AletheIA + Adaptive Skills Integrated Evolution Backlog

Status

• Backlog version: 0.4
• Baseline date: 2026-07-02
• Canonical backlog owner: AletheIA
• Capability-side partner: Adaptive Skills

Registered sources

v0.4 repository sources

These are repository-native sources, not imported packs. They become planning inputs for the next backlog round, but do not create implementation authority by themselves.

This backlog reconciles the original ten evolution packs, supplemental governance packs, screenshots of already-implemented ok-* folders, and newer P13–P18 sources with the implementation already present in AletheIA and Adaptive Skills. It is an executive map and dependency graph, not a replacement for the source packs or canonical contracts.

Naming and numbering normalization

The attached folder screenshots are planning aids, not canonical numbering. This backlog normalizes by capability name and source checksum:

• folders prefixed ok- are treated as delivered or absorbed baseline unless new evidence reopens them;
• duplicate Work Observatory entries remain one canonical P10/P12 line and one S18 evidence gate;
• Design System Intelligence v0.2 supersedes the earlier Pulso Lab package for future planning;
• Runtime 2.0 is a north-star/boundary-review source, not an implementation authorization.

Before implementing an epic, reopen its complete pack. Ask for a new copy only when the archive is unavailable or a newer revision exists.

Operating boundary

Status vocabulary

No requirement may disappear silently. A changed status requires evidence or a decision reference.

Pack reconciliation

Dependency graph

S0–S6 Delivered foundations
  -> S13 Backlog v0.2 + supplemental source registry + M1 north-star map
      -> S7 Bounded debugging pilot (P9)
          -> S14 First Work Observatory dogfood record (P10)

S13
  -> S15 SYSTEM_STATE + Restart Package continuity reconciliation (P11)
      -> S16 Cognitive, intent and documentation closure integration (P11; after S8)
          -> S17 First-use and explainable-language pilot (P11)

S14 + at least five comparable records in one comparison group
  -> S18 Comparative work metrics and read-only projection review (P10/P12)

S17
  -> S19 Backlog v0.3 source registry and prioritization (P13-P18)
      -> S20 Lean Skill Doctrine + Skill Catalog Governance (P17)
          -> S22 Lean Implementation Skill (P17)
      -> S21 Independent Validation Hardening (P16)
          -> S23 Governed Loop Engineering Addendum reconciliation (P14)
      -> S24 Design System Intelligence / Pulso pilot (P15; after S10)
      -> S25 Human Expertise & Evidence-Based Learning (P13; after S20/S21)
      -> S26 Runtime 2.0 boundary review (P18; strategic/deferred)

Existing branches remain:
S1 -> S2 Reference Intake minimum protocol (P2)
S1 -> S8 Intent-to-Evidence contracts (P4) -> S9 Coding Safety specialization (P0)
S1 -> S10 Capability routing reconciliation (P5) -> S11/S24 System-aware Design System pilot (P6/P15)
S4 + S5 + S10 -> S12 Agent-role reconciliation pilot (P7)

S0-S26 closure + S18/S26 blockers
  -> S27 Backlog v0.4 source registry and prioritization
      -> S28 AI Agent Security & Prompt Injection domain-pack minimum (P20)
          -> S29 Web App Security & Trust Boundaries domain-pack minimum (P21)
      -> S30 Enterprise / constrained adoption evidence loop (P22)
      -> S31 Resource-aware next-signals review (P23; does not activate S18)
      -> S32 Observatory maturity review (P1/P3/P10/P11; only after source-backed evidence)
          -> S33 Backlog v0.5 readiness review
              -> S34-S39 Domain Governance usage/evidence cycle
                  -> S40 Backlog v0.5 framing review
                      -> S41 Documentation and first-use coherence review
                          -> S42 Backlog v0.6 opportunity inventory
                              -> S43 Roadmap documentation reference integrity repair
                                  -> S44 Post-S43 backlog gate review
                                      -> S45 Open Dependabot trust-boundary review
                                          -> S46 Dependabot cycle closure reconciliation
                                              -> S47 Backlog development measurement readiness
                                                  -> S48 Backlog development descriptive snapshot
                                                      -> S49 Next-source backlog gate
                                                          -> S50 Learning Distillation & Context Hygiene Addendum v0.7 intake review
                                                              -> S51 Restart Package learning and context hygiene compatibility review
                                                                  -> S52 AletheIA to Adaptive Skills evolution bridge compatibility review
                                                                      -> S53 Minimum sufficient solution guardrail compatibility review
                                                                          -> S54 Observatory interface refinement review

The post-S54 documentation path through S82 is delivered. The new source-triggered continuity path is separate and conditional:

S82
  -> S83 Portable Continuity Reference Intake
      -> intake gate: S84 only
          -> S84 Restart Package portability compatibility review
              -> compatibility gate: absorb/stop or authorize S85
                  -> S85 Codex-to-Codex restart pilot
                      -> S86 same-runtime value decision
                          -> value gate: stop or conditionally authorize S87-S89

The user-observed finalization-feedback gap is a separate, documentation-only path:

S86 closure evidence
  -> S90 User-Facing Closeout Profile
      -> document plain-language completion, validation, publication, continuity and archive-readiness feedback

S8 is delivered and unlocks S9 and the delivered S16 path. S19 is the next safe organizational slice because it reconciles new source packs before any implementation. S20 and S21 should precede new skill creation, broader loops, agents or runtime surfaces. S18 remains deferred until at least five reviewed records share one stable comparison group. P1 remains the delivered visual baseline; P10/P12, P13 and P18 do not reopen live importers, runtime collection, learning automation or new dashboards through this backlog update.

S27 opens backlog v0.4 after S0-S26 closure. Its purpose is to choose the next evidence-respecting track, not to reopen blocked runtime, benchmark, auto-routing, learning automation or dashboard work. Domain governance is the strongest next path because it can improve agent/tool safety and trust-boundary clarity without requiring runtime implementation.

PR-ready slices

S27 — Backlog v0.4 source registry and prioritization

• Repository / layer: AletheIA canonical backlog
• Covers: P19-P23 planning sources and post-S0-S26 prioritization
• Dependencies: S0-S26 delivered/deferred closure; S18 comparison-readiness review; ADR-016 Runtime 2.0 boundary review
• Status: delivered by AletheIA PR #315
• Non-goals: implementing domain packs, creating runtime enforcement, creating dashboards, activating S18, building Runtime 2.0 or changing Adaptive Skills.
• Acceptance: v0.4 sources are registered, next slices are sequenced, blocked tracks remain blocked and every candidate has repository/layer, dependencies, non-goals and acceptance direction.
• Evidence: AletheIA PR #315 / 685b1ab; backlog v0.4 source registry, S27-S32 sequencing, SYSTEM_STATE update and governance/test validation.

S28 — AI Agent Security & Prompt Injection domain-pack minimum

• Repository / layer: AletheIA domain governance pack; optional Adaptive Skills compatibility later
• Covers: P20 instruction trust hierarchy, untrusted content, prompt-injection resistance, tool least privilege, memory/retrieval safety and safe failure.
• Dependencies: S27; existing execution boundaries, AHC/AHGE, restricted knowledge policy, source precedence and tool-risk taxonomy.
• Status: delivered by AletheIA PR #316
• Non-goals: provider-specific jailbreak recipe, technical enforcement engine, secret policy replacement, runtime scanner, or claims of perfect defense.
• Acceptance: one reusable domain-pack contract or guide defines trust categories, escalation boundaries, source refs and adversarial evaluation scenarios without becoming core framework law.
• Evidence: AletheIA PR #316 / a8c492b; ai-agent-security-prompt-injection.md, ai-agent-security-review-checklist.md, ai-agent-security-review-template.yaml, prompt-injection-review.yaml and contract tests.

S29 — Web App Security & Trust Boundaries domain-pack minimum

• Repository / layer: AletheIA domain governance pack
• Covers: P21 client/server separation, secrets, auth/data isolation, tenant-scoped retrieval, authoritative business logic, webhook/integration trust and artifact hygiene.
• Dependencies: S27; S28 recommended if the implementation involves agents consuming web or monitored content.
• Status: delivered by AletheIA PR #318
• Non-goals: vendor tutorial, complete AppSec manual, framework-core expansion, or executable enforcement.
• Acceptance: one reusable web/API trust-boundary pack with review checklist and at least one evaluation scenario.
• Evidence: AletheIA PR #318 / b864419; web-app-security-trust-boundaries.md, web-app-security-review-checklist.md, web-app-security-review-template.yaml, trust-boundary-review.yaml and contract tests.

S30 — Enterprise / constrained adoption evidence loop

• Repository / layer: AletheIA adoption guidance and project-extension examples
• Covers: P22 constrained adoption, local trust-boundary posture, heavier review mapping and project-local extension evidence.
• Dependencies: existing enterprise-readiness roadmap, local trust-boundary posture and constrained adoption examples.
• Status: delivered by AletheIA PR #320
• Non-goals: enterprise product packaging, universal compliance preset, CLI/bootstrap promise or org-specific policy hardcoding.
• Acceptance: one bounded adoption lane maps task brief, gates, handoff, validation and durable decisions to a constrained/local approval context.
• Evidence: AletheIA PR #320 / 0a46aa3; constrained-adoption-evidence-loop-template.yaml, s30-constrained-adoption-evidence-loop.yaml and contract tests, with reusable learning and local-only rules clearly separated.

S31 — Resource-aware next-signals review

• Repository / layer: AletheIA resource/work evidence review
• Covers: P23 next-signal watch-list and S18 blocker monitoring.
• Dependencies: S27 and current Work Observatory comparison-readiness review.
• Status: delivered by AletheIA PR #322
• Non-goals: activating S18, benchmark packaging, vendor ranking, auto-routing, learning-layer behavior or new telemetry collection.
• Acceptance: review accumulated evidence for repeated comparable slices, late-stage waste, local rule repetition and stable reinforced outcomes; record whether thresholds remain unmet.
• Evidence: AletheIA PR #322 / ef25f60; s31-next-signals-review.json reviews current Work Observatory records and keeps S18 deferred; unavailable inputs remain unavailable.

S32 — Observatory maturity review

• Repository / layer: AletheIA Visual/Resource/Work Observatory planning
• Covers: P1, P3, P10/P12 and P11 deferred visual/evidence projections.
• Dependencies: S31 or a new source-backed evidence review showing repeated need.
• Status: delivered by AletheIA PR #324
• Non-goals: live importers, collectors, dashboards, ranking, people scoring or automatic documentation-health inference.
• Acceptance: decide whether existing Observatory surfaces need refinement, not expansion by default; every visual claim must have source_refs and unavailable state.
• Evidence: AletheIA PR #324 / b93b3d7; s32-observatory-maturity-review.json records the maturity decision; no-build/refine-when-needed outcome remains accepted.

S33 — Backlog v0.5 readiness review

• Repository / layer: AletheIA backlog planning
• Covers: post-S27-S32 cycle closure, P19 partial-delivery posture and deferred-track readiness.
• Dependencies: S27-S32 delivered and v0.4 traceability refreshed.
• Status: delivered by AletheIA PR #327
• Non-goals: activating S18, Runtime 2.0, collectors, dashboards, policy engines, automatic routing, new domain packs or Adaptive Skills changes.
• Acceptance: record a source-backed next-cycle posture, preserve blocked/deferred tracks, and select the next safe candidate without expanding implementation scope.
• Evidence: AletheIA PR #327 / 5dfd0a4; s33-v05-readiness-review.json selects a domain-governance usage review as the next safe candidate and keeps S18, Runtime 2.0 and Observatory expansion deferred.

S34 — Domain Governance Usage Review

• Repository / layer: AletheIA domain governance planning
• Covers: P19 domain-pack usage posture after S28/S29 delivery.
• Dependencies: S28, S29, S33 and v0.4 traceability refresh.
• Status: delivered by AletheIA PR #329
• Non-goals: new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes or framework-core authority changes.
• Acceptance: review whether delivered domain packs are usable as advisory guidance, record gaps as unavailable where evidence is missing, and define source-backed next signals before expansion.
• Evidence: AletheIA PR #329 / f3ccc9f; s34-domain-governance-usage-review.json keeps S28/S29 usable as manual/advisory guidance and defers new packs or enforcement until repeated usage evidence exists.

S35 — Domain Governance Usage Evidence Capture

• Repository / layer: AletheIA domain governance evidence capture
• Covers: future P19/S28/S29 usage evidence capture before expansion.
• Dependencies: S34 delivered.
• Status: delivered by AletheIA PR #331
• Non-goals: claiming real usage evidence, new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes or S18 activation.
• Acceptance: add a minimal template and reviewed example that define how future real usage evidence is captured while preserving unavailable values and confirming the template itself is not evidence.
• Evidence: AletheIA PR #331 / 256436d; domain-governance-usage-evidence-template.yaml and s35-domain-governance-usage-evidence-capture.json define future capture without claiming usage.

S36 — Dependabot Trust-Boundary Usage Review

• Repository / layer: AletheIA domain governance usage evidence
• Covers: first real/source-backed S28/S29 usage evidence from the completed Dependabot dependency PR cycle.
• Dependencies: S35 capture template and merged dependency PRs #273, #293, #292, #291 and #290.
• Status: delivered by AletheIA PR #333
• Non-goals: new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes, S18 activation or security/success-rate claims.
• Acceptance: apply S28/S29 as manual/advisory review guidance to the real dependency-maintenance cycle, capture evidence with the S35 shape, and keep expansion blocked until repeated real usage evidence exists.
• Evidence: AletheIA PR #333 / 91c2a55; s36-dependabot-trust-boundary-review.json records Dependabot PR usage evidence while keeping the packs advisory and non-enforcing.

S37 — Domain Governance Next-Use Readiness Review

• Repository / layer: AletheIA domain governance readiness
• Covers: criteria for when a future real Work Slice should apply S28/S29 and capture evidence with the S35 template.
• Dependencies: S35 capture template and S36 delivered first real usage record.
• Status: delivered by AletheIA PR #335
• Non-goals: claiming new real usage evidence, new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes, S18 activation or security/success-rate claims.
• Acceptance: define source-backed entry criteria, candidate triggers and non-triggers for future S28/S29 capture while confirming readiness planning does not count as usage evidence.
• Evidence: AletheIA PR #335 / 38dabc0; s37-domain-governance-next-use-readiness.json keeps the next-use gate explicit and blocked until a real Work Slice appears.

S38 — Adaptive Skill Record Trust-Boundary Review

• Repository / layer: AletheIA Mission Control domain governance usage evidence
• Covers: real S28/S29 usage on the Adaptive Skills execution-record adapter boundary in Mission Control.
• Dependencies: S35 capture template, S37 readiness criteria and the existing Mission Control Adaptive Skills execution-record adapter.
• Status: delivered by AletheIA PR #337
• Non-goals: new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes, S18 activation, auth/tenant/secrets changes or security/success-rate claims.
• Acceptance: apply S28/S29 to the real cross-repository record projection boundary, add a focused regression test for prompt-like record content, and capture source-backed usage evidence without expanding governance.
• Evidence: AletheIA PR #337 / 1ea0725; s38-adaptive-skill-record-trust-boundary-review.json records the S28/S29 usage decision and adapter test coverage.

S39 — Domain Governance Evidence Sufficiency Review

• Repository / layer: AletheIA domain governance sufficiency review
• Covers: whether S36 and S38 provide enough real S28/S29 usage evidence to justify domain-governance expansion.
• Dependencies: S35 capture template, S36 and S38 real usage records, and S37 readiness criteria.
• Status: delivered by AletheIA PR #339
• Non-goals: new domain packs, scanners, policy engines, runtime enforcement, dashboards, schemas, automatic classification, Adaptive Skills changes, S18 activation, auth/tenant/secrets changes or security/success-rate claims.
• Acceptance: compare the real usage records, separate readiness-only S37 from usage evidence, and decide whether evidence is sufficient for refinement or expansion.
• Evidence: AletheIA PR #339 / ca36bab; s39-domain-governance-evidence-sufficiency-review.json keeps expansion blocked because S36/S38 are useful but still too few and heterogeneous.

S40 — Backlog v0.5 framing review

• Repository / layer: AletheIA backlog planning
• Covers: post-S34-S39 cycle closure, S28/S29 usage posture, deferred-track blockers and next safe v0.5 candidate selection.
• Dependencies: S33 readiness review, S35 capture template, S37 next-use gate and S39 sufficiency review.
• Status: delivered by AletheIA PR #341
• Non-goals: S18 activation, Runtime 2.0 implementation, collectors, dashboards, scanners, policy engines, enforcement, schemas, new domain packs, automatic classification, Adaptive Skills changes or treating synthetic/template examples as real evidence.
• Acceptance: record the source-backed post-S39 posture, keep domain-governance expansion blocked, preserve deferred tracks and select the next safe backlog candidate without introducing implementation authority.
• Evidence: AletheIA PR #341 / a256365; s40-backlog-v05-framing-review.json closes the S34-S39 usage/evidence cycle and recommends a bounded documentation and first-use coherence review next.

S41 — Documentation and first-use coherence review

• Repository / layer: AletheIA reader-facing documentation and backlog evidence
• Covers: source-backed coherence of README entry points, first-use guidance, core operating path, SYSTEM_STATE posture and the changelog’s release-history role.
• Dependencies: S40 backlog v0.5 framing review.
• Status: delivered by AletheIA PR #346.
• Non-goals: automatic documentation generation, a new onboarding product UI, mandatory ceremony for Lite slices, S18 activation, Runtime 2.0 implementation, collectors, dashboards, scanners, policy engines, enforcement, schemas, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: identify reader-facing coherence gaps from the named sources; correct only bounded, source-backed documentation blockers; keep current state and release-history surfaces distinguishable.
• Evidence: AletheIA PR #346 / 679bb6c; s41-documentation-first-use-coherence-review.json records the scope, findings and bounded README path corrections.

S42 — Backlog v0.6 opportunity inventory

• Repository / layer: AletheIA backlog planning and source-backed documentation inventory
• Covers: the post-S41 candidate set, including evidence-gated tracks, real-work triggers and observable reader-facing documentation gaps in current roadmap sources.
• Dependencies: S40 framing review and S41 documentation/first-use coherence review.
• Status: delivered by AletheIA PR #348.
• Non-goals: treating planning as a real S28/S29 usage record, activating S18, implementing Runtime 2.0, creating collectors, dashboards, scanners, policy engines, enforcement, schemas, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: inventory the candidates with their evidence posture; keep gated tracks gated; select at most one small source-backed next move without creating implementation authority.
• Evidence: AletheIA PR #348 / 4734661; s42-backlog-v06-opportunity-inventory.json records the candidate decision.

S43 — Roadmap documentation reference integrity repair

• Repository / layer: AletheIA roadmap documentation
• Covers: repair of the unresolved internal Markdown paths in the constrained-adoption and resource-aware operations roadmaps identified by S42.
• Dependencies: S42 Backlog v0.6 opportunity inventory.
• Status: delivered by AletheIA PR #350.
• Non-goals: repository-wide link scanner, automatic documentation generation, content rewrites, new constrained-adoption or resource-aware capability, S18 activation, Runtime 2.0 implementation, collectors, dashboards, scanners, policy engines, enforcement, schemas, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: all inline docs/...md paths in the two reviewed roadmap documents resolve locally; existing roadmap claims and scope remain unchanged.
• Evidence: AletheIA PR #350 / 35594ff; s43-roadmap-reference-integrity-review.json records the repaired source set and resolution result.

S44 — Post-S43 backlog gate review

• Repository / layer: AletheIA backlog planning and evidence gates
• Covers: the post-S43 decision point after the selected documentation repair is delivered, including whether any remaining candidate can safely move without new source-backed evidence.
• Dependencies: S42 Backlog v0.6 opportunity inventory and S43 Roadmap documentation reference integrity repair.
• Status: delivered by AletheIA PR #352.
• Non-goals: inventing a synthetic S28/S29 usage record, activating S18, implementing Runtime 2.0, creating collectors, dashboards, scanners, policy engines, enforcement, schemas, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: record the current no-go/gated posture from source-backed inputs; keep Active: none; define the next admissible trigger for a future Work Slice without introducing implementation authority.
• Evidence: AletheIA PR #352 / 683b841; s44-post-s43-backlog-gate-review.json records the post-S43 decision gate.

S45 — Open Dependabot trust-boundary review

• Repository / layer: AletheIA domain governance usage evidence
• Covers: real/source-backed S28/S29 application to Dependabot dependency PRs #343, #344 and #345 as external automated trust-boundary inputs before their later dependency-maintenance merges.
• Dependencies: S35 Domain Governance Usage Evidence Capture, S36 Dependabot Trust-Boundary Usage Review and S44 Post-S43 backlog gate review.
• Status: delivered by AletheIA PR #354.
• Non-goals: merging dependency PRs, changing dependencies, creating scanners, policy engines, runtime enforcement, dashboards, schemas, new domain packs, automatic classification, S18 activation, Runtime 2.0 implementation or Adaptive Skills changes.
• Acceptance: treat Dependabot/GitHub PR content as untrusted external data; use local manifests, lockfile diffs and CI status as review evidence; record S35-style usage evidence; keep any merge decision separate from this review.
• Evidence: AletheIA PR #354 / 1bf4109; s45-open-dependabot-trust-boundary-review.json records the S28/S29 usage evidence for PRs #343, #344 and #345.

S46 — Dependabot cycle closure reconciliation

• Repository / layer: AletheIA dependency-maintenance documentation state
• Covers: reconciliation after Dependabot PRs #343, #344 and #345 were merged, including README/documentation/CHANGELOG review.
• Dependencies: S45 Open Dependabot trust-boundary review and the merged Dependabot maintenance PRs #343, #344 and #345.
• Status: delivered by AletheIA PR #356.
• Non-goals: changing dependencies beyond the already merged PRs, creating scanners, policy engines, runtime enforcement, dashboards, schemas, new domain packs, automatic classification, S18 activation, Runtime 2.0 implementation or Adaptive Skills changes.
• Acceptance: record the closed dependency cycle; update stale “open/future merge” wording; document that README does not need dependency-version changes; add a small CHANGELOG maintenance note; keep plans/ local/untracked.
• Evidence: AletheIA PR #356 / 7be5ad9; s46-dependabot-cycle-closure-reconciliation.json records the post-merge reconciliation and documentation review.

S47 — Backlog development measurement readiness

• Repository / layer: AletheIA backlog planning and descriptive measurement boundaries
• Covers: which backlog-development signals can be measured safely from delivered PR/work-slice history without activating S18 or creating comparative metrics.
• Dependencies: S18 comparison-readiness review, S44 post-S43 gate, S45/S46 Dependabot cycle evidence and current SYSTEM_STATE posture.
• Status: delivered by AletheIA PR #358.
• Non-goals: S18 activation, success rates, productivity ranking, value/cost scoring, work-unit calibration, dashboards, collectors, scanners, policy engines, runtime enforcement, schemas, Runtime 2.0 implementation, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: separate descriptive operational counters from inadmissible comparative claims; define a minimal future record shape for backlog-development cycles; confirm that this supports continuing backlog development rather than pausing work.
• Evidence: AletheIA PR #358 / 0d2ba32; s47-backlog-development-measurement-readiness.json records the measurement boundary and next-safe usage.

S48 — Backlog development descriptive snapshot

• Repository / layer: AletheIA backlog planning and descriptive measurement record
• Covers: a first S47-compliant descriptive snapshot of the recent S40-S47 backlog-development cycle.
• Dependencies: S47 Backlog development measurement readiness and delivered S40-S47 backlog slices.
• Status: delivered by AletheIA PR #360.
• Non-goals: S18 activation, success rates, productivity ranking, value/cost scoring, work-unit calibration, dashboards, collectors, scanners, policy engines, runtime enforcement, schemas, Runtime 2.0 implementation, new domain packs, automatic classification or Adaptive Skills changes.
• Acceptance: record only descriptive counts and source refs for the cycle; keep comparison group unavailable; confirm no success-rate, ranking or efficiency claim is derived.
• Evidence: AletheIA PR #360 / ae6bb7b; s48-backlog-development-descriptive-snapshot.json records the bounded descriptive snapshot.

S49 — Next-source backlog gate

• Repository / layer: AletheIA backlog planning and source-intake boundary
• Covers: criteria for what may safely reopen the next backlog round after S48 without inventing implementation.
• Dependencies: S47 Backlog development measurement readiness, S48 Backlog development descriptive snapshot, S35/S37/S39 domain-governance evidence boundaries and current Active: none posture.
• Status: delivered by AletheIA PR #362.
• Non-goals: S18 activation, success rates, productivity ranking, value/cost scoring, work-unit calibration, dashboards, collectors, scanners, policy engines, runtime enforcement, schemas, Runtime 2.0 implementation, new domain packs, automatic classification, Adaptive Skills changes or treating planning/templates as real usage evidence.
• Acceptance: define admissible source kinds, non-triggers and required handling for the next backlog round; keep future development unblocked only when a real source appears.
• Evidence: AletheIA PR #362 / 56b4113; s49-next-source-gate.json records the next-source gate and preserves existing blockers.

S0 — Integrate skill execution context

• Repository / layer: AletheIA / Resource Observatory
• Covers: P1 operational baseline; P9 prerequisite
• Dependencies: none
• Status: delivered by PR #250
• Acceptance: mode, result, modules, handoff and evidence remain optional, source-backed and inspector-only.
• Evidence: merge commit d7c35bf; CI governance, build, tests and visual snapshots passed.

S1 — Publish integrated backlog and traceability

• Repository / layer: AletheIA canonical; Adaptive Skills cross-link
• Covers: all packs
• Dependencies: S0
• Non-goals: importing pack files into canonical contracts; implementing any pack wholesale.
• Acceptance: every pack has posture, requirement-family mapping, dependencies, acceptance evidence and deferred/rejected rationale.
• Evidence: governance check, link validation and clean diff.

S2 — Reference Intake minimum protocol

• Repository / layer: AletheIA governance; Adaptive Skills capability-side proposal template
• Covers: P2
• Dependencies: S1
• Status: delivered through AletheIA PR #297
• Non-goals: dependency installation, automatic adoption, runtime authority changes or skill mutation.
• Acceptance: one intake contract, one adoption decision template and one worked example based on this archive.
• Evidence: AletheIA PR #297 / 53b555a; reference-intake-adoption-contract.md, reference-intake-decision-template.yaml, evolution-pack-intake.yaml and guardrail tests make source provenance, layer mapping and adopt/adapt/reference/defer/reject decisions explicit.

S3 — Pattern context unavailable states

• Repository / layer: AletheIA / Resource Observatory view model
• Covers: P9 pattern selection and compatibility fallback
• Dependencies: S0, S1
• Status: delivered by PR #252
• Non-goals: new page, card, persistence schema, selection action or loop execution.
• Acceptance: optional patternContext renders no-context, unavailable compatibility and missing-gate states without inference.
• Evidence: merge commit 0d8f8b8; typed fixtures, adapter/inspector tests, keyboard/focus regression, desktop/mobile browser QA and CI snapshots.

S4 — Context Surface minimum registry

• Repository / layer: AletheIA context governance
• Covers: P8
• Dependencies: S1
• Status: delivered by AletheIA PR #254 and Adaptive Skills PR #65
• Non-goals: automatic context router or mandatory registry of every document.
• Acceptance: each registered surface declares owner, scope, load mode, path, activation, freshness, budget and allowed/prohibited content.
• Evidence: AletheIA merge 3f2df81 provides the contract, template and Work Slice example; Adaptive Skills merge 3606089 provides the capability-scoped provider-loading example; both repositories passed their full CI gates.

S5 — Recoverable observation contract

• Repository / layer: AletheIA contract; Adaptive Skills return pattern
• Covers: P3
• Dependencies: S1; aligns with S4
• Status: delivered by AletheIA PR #256 and Adaptive Skills PR #66
• Non-goals: terminal proxy, RTK dependency, evidence database or automatic normalizer.
• Acceptance: structured observations preserve decision-relevant evidence; any lossy observation requires a recovery pointer; unavailable metrics are not invented.
• Evidence: AletheIA merge a5f3e9c provides the contract, template and test-output example; Adaptive Skills merge 60de75d provides the portable skill return, template and feature-planning example; both repositories passed their full CI gates.

S6 — Source-backed pattern context

• Repository / layer: AletheIA projection adapters; Adaptive Skills compatibility source
• Covers: P9 observe milestone
• Dependencies: S3, S5; canonical execution-pattern records already exist
• Status: delivered by AletheIA PR #258
• Non-goals: pattern selection or compatibility inference inside the UI.
• Acceptance: inspector separates pattern, mode, autonomy and result; every displayed field carries a source reference.
• Evidence: merge ddb6b32 adds a read-only projection adapter, a schema-valid debugging selection fixture, canonical Adaptive Skills declaration refs, and compatible, incompatible, unavailable, missing-objective-gate and analytics-gating tests; full CI passed.

S7 — Bounded debugging loop pilot

• Repository / layer: AletheIA governance; Adaptive Skills debugging declaration; runtime evidence
• Covers: P9 pilot milestone
• Dependencies: S5, S6
• Status: delivered by AletheIA PR #262
• Non-goals: generic loop engine, strategic loops, automatic merge or skill self-update.
• Acceptance: one reproducible failure; loop_until_done; maximum three iterations; explicit budget; objective test/reproduction gate; evidence per iteration; human review before merge.
• Evidence: merge a6acc4f preserves absent iteration counts as unavailable, records the debugging@0.1.0 activation, Codex executor, no subagents, invoked tools, one passing iteration, explicit budgets, browser validation and the accepted human review in PR #262.

S8 — Intent-to-Evidence minimum contracts

• Repository / layer: AletheIA Work Slice; Adaptive Skills consultative pattern
• Covers: P4
• Dependencies: S1
• Status: delivered by AletheIA PR #268 and Adaptive Skills PR #70; human review accepted before merge.
• Non-goals: replacing PRD/SDD, mandatory use for Lite slices or agent-authored intent without confirmation.
• Acceptance: optional Intent and Expectations references, guessing-risk review and evidence-to-expectation map integrate with readiness and reconcile.
• Evidence: AletheIA merge 2575e07 adds the optional extension, templates, synthetic example, real bounded pilot and tests; Adaptive Skills merge 59c0f41 adds the advise-only intent-clarification capability, template, example, harness declaration and registries. The project owner accepted the intent and expectations before both PRs were merged.

S9 — Coding Safety specialization

• Repository / layer: AletheIA coding plan profile; Adaptive Skills planning advisor candidate
• Covers: P0
• Dependencies: S8
• Status: delivered through AletheIA PR #299
• Non-goals: new runtime, /improve command, autonomous coding, policy engine, dependency installation, schema or auto-merge.
• Acceptance: plan declares base state, in/out scope, verification commands and expected results, drift check, rollback and stop conditions.
• Evidence: AletheIA PR #299 / 9579df5; coding-safety-plan-profile.md, coding-safety-plan-template.yaml, coding-safety-non-engineer-checklist.md, s9-standard-docs-slice.yaml and guardrail tests.

Promote implementation-planning-advisor only if the pilot proves a recurring gap not covered by feature-planning, testing, debugging, architecture-review, Lean Implementation and independent validation guidance.

S10 — Capability routing reconciliation

• Repository / layer: AletheIA governance; Adaptive Skills capability metadata
• Covers: P5
• Dependencies: S1
• Status: delivered by AletheIA PR #276 and Adaptive Skills PR #77.
• Non-goals: routing engine, hard-coded provider selection or global routing owned by Adaptive Skills.
• Acceptance: one canonical mapping explains tool, skill, subagent, execution vehicle, execution pattern and AHC without adding competing terms.
• Evidence: AletheIA merge 393cab4 adds docs/contracts/capability-routing-reconciliation.md, four worked decisions, Observatory projection boundaries and tests; Adaptive Skills merge 156dc9c adds the capability routing boundary note.

S11 — System-aware Design System pilot

• Repository / layer: Adaptive Skills capability module; AletheIA pilot governance
• Covers: first bounded slice of P6
• Dependencies: S2, S4, S10
• Status: delivered through S24 / AletheIA PR #287 and Adaptive Skills PR #81
• Non-goals: adopting the entire SAGP umbrella or creating many first-class skills.
• Acceptance: one Pulso-aware review loads the governing Design System only when relevant, reports compliance and exceptions, and creates an evolution proposal only when evidence supports it.
• Evidence: S24 delivered the same bounded Pulso-aware pilot through the AletheIA Design System Intelligence pilot contract, starter-pack template, synthetic Pulso review, Adaptive Skills design-system-intelligence skill, capability metadata and validation commands. This closes the original S11 pilot slot without promoting the broader System-Aware Governance umbrella.

S12 — Agent-role contract reconciliation pilot

• Repository / layer: AletheIA roles and adapters; Adaptive Skills provider integration remains future repository-side alignment
• Covers: P7
• Dependencies: S4, S5, S10
• Status: delivered through AletheIA PR #302
• Non-goals: seven autonomous agents, provider lock-in, duplicated role catalog, routing engine, runtime integration or Adaptive Skills mutation.
• Acceptance: existing Software Engineer and QA/Governance Reviewer roles declare authority, allowed capabilities/tools, evidence, stop and handoff behavior for one real Work Slice.
• Evidence: AletheIA PR #302 / da75d60; agent-role-contract-reconciliation.md, agent-role-plan-template.yaml, s12-software-engineer-qa-governance.yaml and guardrail tests provide a role plan, review record, closure brief and provider references.

S13 — Register supplemental packs and north-star map

• Repository / layer: AletheIA canonical backlog and concepts; Adaptive Skills cross-link
• Covers: P10, P11 and M1
• Dependencies: S1; current repository evidence through S6
• Status: delivered by AletheIA PR #260 and Adaptive Skills PR #67
• Non-goals: implementing either pack, promoting the map to architecture authority, creating ADRs, schemas, runtime components or visual metrics.
• Acceptance: all sources have checksums; P10/P11 have postures, requirement mapping, dependencies and incremental slices; M1 is explicitly non-normative and illustrative.
• Evidence: AletheIA merge 4b08b0f provides the versioned map, concept note, v0.2 traceability and indexes; Adaptive Skills merge feafc03 provides the canonical cross-link; both repositories passed their applicable CI gates and plans/ remained untouched.

S14 — First Work Observatory dogfood record

• Repository / layer: AletheIA derived work observation; Adaptive Skills return reference when used
• Covers: P10 minimum record
• Dependencies: S7 and S13
• Status: delivered by AletheIA PR #264
• Non-goals: work-unit scoring, value scoring, people evaluation, database, collector, dashboard or comparative claims.
• Acceptance: S7 produces one source-backed work_record with duration, outcome, execution/resource/observation refs, human intervention and provenance; acceptance and rework remain unknown until reviewed.
• Evidence: merge 5835670 adds the first metadata-only derived record and guardrail tests for S7. Later S12, S21, S23 and S25 dogfood records add source-backed work observations for agent-role reconciliation, independent validation hardening, loop readiness and human-expertise learning. All keep unsupported duration, resource, acceptance, rework, work-unit, quality and value claims explicitly unavailable or unknown.

S15 — SYSTEM_STATE and continuity reconciliation

• Repository / layer: one small local state registry per repository; AletheIA continuity governance
• Covers: P11 continuity and project-state baseline
• Dependencies: S13
• Status: delivered by AletheIA PR #266 and Adaptive Skills PR #68
• Non-goals: universal truth database, automatic documentation generator, monolithic always-loaded prompt or duplicate Continuity Capsule lifecycle.
• Acceptance: AletheIA and Adaptive Skills each maintain a compact first-load SYSTEM_STATE; Continuity Capsule fields extend the existing Restart Package and post-resume check; contracts, ADRs and evidence remain authoritative.
• Evidence: AletheIA merge ec9e9af adds the state contract, root index, Restart Package compatibility fields, post-resume check, context-surface entries and tests; Adaptive Skills merge baf4a4a adds its local state, context entry, validator and required CI gate.

S16 — Cognitive, intent and documentation closure integration

• Repository / layer: AletheIA Work Slice closure governance
• Covers: P11 cognitive debt, intent debt and documentation coherence
• Dependencies: S8 and S15
• Status: delivered by AletheIA PR #270 after project-owner review.
• Non-goals: blocking every Lite slice, agent self-certification, automatic documentation mutation or replacing technical/security review.
• Acceptance: proportional closure review reuses comprehension debt and durable decisions, identifies required documentation updates and escalates high-risk understanding gaps without creating parallel lifecycle authority.
• Evidence: merge fb6faba adds the proportional closure contract, finalization integration, low-risk docs case, stale-documentation case, high-risk human-review case, source-backed real closure brief and five guardrail tests.

S17 — First-use and explainable-language pilot

• Repository / layer: AletheIA guidance; Adaptive Skills usage documentation
• Covers: P11 language modes, cognitive boundary and first-run guidance
• Dependencies: S16
• Status: delivered by AletheIA PR #272 and Adaptive Skills PR #73 after project-owner review.
• Non-goals: hiding technical risk, maintaining four duplicate guide sets or auto-detecting user expertise without confirmation.
• Acceptance: one low-risk first-use journey supports plain/guided/professional/expert depth, expected outputs, validation, troubleshooting and explicit technical-escalation boundaries.
• Evidence: AletheIA merge 6dba175 supplies the single adaptive-depth journey, Explainable Change Brief and tests; Adaptive Skills merge e4a5591 aligns safe skill first use. Guided and expert fixtures preserve equivalent validation/escalation, with a bounded readability review.

S18 — Comparative work metrics and projection review

• Repository / layer: Work Observatory derived metrics; existing Mission Control/Resource Observatory surface if activated
• Covers: P10 comparative and visual milestones
• Dependencies: S14 plus at least five reviewed records in one stable comparison_group
• Status: deferred until the evidence threshold is met; PR #313 adds a comparison-readiness review confirming the current five records are not eligible.
• Non-goals: new page by default, work-unit leaderboards, individual productivity scoring, fake ROI or mixed-task comparisons.
• Acceptance: only source-backed metrics with formula, scope, provenance, confidence and limitations are eligible; unavailable inputs suppress dependent metrics; visual projection remains read-only.
• Evidence: comparison-readiness-review.json records the current denominator audit and keeps S18 blocked; future activation still requires a comparison-validity report, privacy review and explicit activation decision.

S19 — Backlog v0.3 source registry and prioritization

• Repository / layer: AletheIA canonical backlog
• Covers: P13–P18 plus screenshot-based implemented/absorbed plan normalization
• Dependencies: S17 delivered baseline; S10 delivered reconciliation baseline
• Status: delivered by AletheIA PR #278.
• Non-goals: implementing any pack, creating runtime surfaces, creating Adaptive Skills changes or treating screenshots as canonical evidence.
• Acceptance: new source packs have checksums, posture, destination, dependency order and deferred/rejected rationale; ok-* folders are preserved as delivered/absorbed baseline; numbering divergence is normalized by capability name.
• Evidence: AletheIA merge bb342fc registers backlog v0.3 source registry, dependency graph and priority ordering.

S20 — Lean Skill Doctrine + Skill Catalog Governance

• Repository / layer: Adaptive Skills catalog governance; AletheIA governance contract
• Covers: P17; public issues AletheIA #274 and Adaptive Skills #74
• Dependencies: S19
• Status: delivered through AletheIA PR #280 and Adaptive Skills PR #78.
• Non-goals: runtime policy engine, automatic skill routing, blocking inside SKILL.md, rewriting existing skills or merging AletheIA governance into skill methods.
• Acceptance: Adaptive Skills defines lean doctrine and Skill Quality Gate; AletheIA defines skill selection/proposal governance, over-selection signals, narrow blocking cases and provenance requirements.
• Evidence: Adaptive Skills doctrine docs and quality-gate template; AletheIA docs/contracts/skill-catalog-governance-contract.md; cross-links and repository validators.

S21 — Independent Validation Hardening

• Repository / layer: AletheIA validation governance and starter-pack review surfaces
• Covers: P16
• Dependencies: S19; aligns with S8, S16, maker-checker and AHC/AHGE
• Status: delivered through AletheIA PR #281.
• Non-goals: making LLM review proof, requiring critic review for every low-risk task, autonomous merge/deploy, benchmark framework or hidden chain-of-thought transfer.
• Acceptance: proportional independent critic context, evidence-to-expectation gate, validation review record and blind-review checklist exist with examples where justified.
• Evidence: AletheIA PR #281 / fb9f900; docs/contracts/independent-validation-hardening-contract.md, docs/reference/independent-validation-blind-review-checklist.md, starter-pack/templates/independent-validation-review-template.yaml, examples/independent-validation/synthetic-review-case.json and contract tests.

S22 — Lean Implementation Skill

• Repository / layer: Adaptive Skills engineering skill; AletheIA observation compatibility
• Covers: P17 lean implementation follow-up; public issue Adaptive Skills #75
• Dependencies: S20 quality gate and catalog doctrine
• Status: delivered through Adaptive Skills PR #79 and AletheIA PR #285.
• Non-goals: replacing architecture review, testing, debugging, security, accessibility or data-integrity requirements; adding benchmark harness in v0.1.
• Acceptance: lean-implementation passes the Skill Quality Gate, keeps 3–5 core moves, preserves safety exceptions and emits a compact observation-compatible result.
• Evidence: Adaptive Skills PR #79 / 95408c4; AletheIA PR #285 / 9bc12f9; Adaptive Skills lean-implementation skill, quality-gate record, projection status, synthetic engineering example and AletheIA observation-compatibility fixture.

S23 — Governed Loop Engineering Addendum reconciliation

• Repository / layer: AletheIA loop governance; Adaptive Skills loop compatibility where needed
• Covers: P14
• Dependencies: S10, S21 and existing ADR-015 / objective-gate / loop-state / maker-checker contracts
• Status: delivered through AletheIA PR #283.
• Non-goals: scheduler, loop runtime, self-running agents, automatic permissions or loops without objective gates.
• Acceptance: one Loop Readiness Gate and review-capacity/drift guidance reconcile with existing loop contracts without duplicating them.
• Evidence: AletheIA PR #283 / 0b0531a; docs/contracts/loop-readiness-gate.md, docs/reference/loop-readiness-review-checklist.md, starter-pack/templates/loop-readiness-gate-template.yaml, examples/governed-loops/s23-loop-readiness-synthetic.json and guardrail tests.

S24 — Design System Intelligence / Pulso pilot

• Repository / layer: Adaptive Skills design-system governance module; AletheIA pilot governance
• Covers: P6 plus P15 v0.2
• Dependencies: S10 and S20; S21 recommended before review-heavy adoption
• Status: delivered through Adaptive Skills PR #81 and AletheIA PR #287.
• Non-goals: making Pulso a structural dependency, institutionalizing patterns from statistics alone, automatic remediation, broad DS automation or many first-class skills.
• Acceptance: one Pulso-aware review uses source registry, conformance/candidate outputs and Pattern Generalization Gate; at least one actionable finding is classified without promotion authority.
• Evidence: Adaptive Skills PR #81 / a7c5dc2; AletheIA PR #287 / e180f0a; AletheIA Design System Intelligence pilot contract, starter-pack template, synthetic Pulso review, Adaptive Skills design-system-intelligence skill, capability metadata and validation commands.

S25 — Human Expertise & Evidence-Based Learning

• Repository / layer: AletheIA human planning and learning governance; Adaptive Skills evidence hints where appropriate
• Covers: P13
• Dependencies: S20 and S21
• Status: delivered through AletheIA PR #289.
• Non-goals: unrestricted autoevolution, skills mutating themselves, replacing human review at high risk, or capability fitness rankings without evidence.
• Acceptance: smallest useful Domain Expertise Brief and Evidence-Based Learning Record capture domain rules, hypotheses, observed outcomes, surprises and future recommendations without creating a learning automation layer.
• Evidence: AletheIA PR #289 / 0800cd0; human expertise and evidence-based learning contract, starter-pack templates, synthetic governed slice and contract tests.

S26 — Runtime 2.0 boundary review

• Repository / layer: AletheIA architecture decision / strategic roadmap
• Covers: P18
• Dependencies: S19; considers outcomes from S20–S25 before implementation
• Status: delivered through ADR-016 / AletheIA PR #295
• Non-goals: runtime kernel, SDK, CLI, plugin interface, provider adapter, capability runtime or execution engine.
• Acceptance: one boundary decision states whether Runtime 2.0 remains a north-star, becomes a future track, or is rejected/absorbed into existing harness/runtime boundaries.
• Evidence: ADR-016 — Runtime 2.0 Boundary Review keeps Runtime 2.0 as a strategic north-star, preserves current AletheIA / Adaptive Skills / runtime-harness boundaries and requires a later explicit implementation decision before code begins.

S50 — Learning Distillation & Context Hygiene Addendum v0.7 intake review

• Repository / layer: AletheIA reference intake and backlog planning
• Covers: P24 learning distillation, context hygiene, Restart Package/finalization extension ideas, AletheIA-to-Adaptive-Skills evolution bridge and minimum-sufficient-solution guardrail.
• Dependencies: S2 reference intake protocol, S49 next-source backlog gate, current Restart Package/SYSTEM_STATE/finalization boundaries and Adaptive Skills governance boundaries.
• Status: delivered by AletheIA PR #365
• Non-goals: implementing learning automation, memory writeback, new runtime behavior, new schemas as canonical contracts, Adaptive Skills changes, collectors, dashboards, policy engines, automatic classification, S18 activation or Runtime 2.0 implementation.
• Acceptance: register the v0.7 package checksum and provenance, mark v0.1-v0.6 variants of the same addendum as superseded, map the source to existing AletheIA/Adaptive Skills boundaries, and record a conservative adoption posture before any implementation proposal.
• Evidence: AletheIA PR #365 / 808420c; s50-learning-distillation-context-hygiene-intake.yaml records the source hash, extracted requirement families, layer mapping, deferred implementation items and guardrails.

S51 — Restart Package learning and context hygiene compatibility review

• Repository / layer: AletheIA finalization/restart guidance
• Covers: P24-R1 learning distillation and context hygiene as a compatibility extension to the existing Restart Package.
• Dependencies: S50 delivered intake, existing slice finalization guidance and Restart Package compatibility fields.
• Status: delivered by AletheIA PR #367
• Non-goals: learning automation, automatic classification, memory writeback, canonical schema adoption, Adaptive Skills mutation, collectors, dashboards, S18 activation or Runtime 2.0 implementation.
• Acceptance: add a manual optional learning/context-hygiene compatibility block to the existing Restart Package guidance and template while preserving Restart Package as the only task-level continuity artifact.
• Evidence: AletheIA PR #367 / 0860531; docs/guides/slice-finalization-and-restart.md, starter-pack/templates/slice-finalization-review-template.md and continuity tests.

S52 — AletheIA to Adaptive Skills evolution bridge compatibility review

• Repository / layer: AletheIA skill evolution governance
• Covers: P24-R2 AletheIA to Adaptive Skills evolution bridge as a compatibility extension to the existing Skill Evolution Validation Contract.
• Dependencies: S50 delivered intake, S51 Restart Package compatibility review and existing Skill Catalog Governance / Skill Evolution Validation contracts.
• Status: delivered by AletheIA PR #369
• Non-goals: changing Adaptive Skills, creating skills, automatic skill mutation, memory writeback, canonical schema adoption, collectors, dashboards, S18 activation or Runtime 2.0 implementation.
• Acceptance: record the manual routing categories and evidence requirements for turning Work Slice learning into an Adaptive Skills evolution signal while preserving proposal/human-review/PR governance.
• Evidence: AletheIA PR #369 / cf60bd9; docs/contracts/skill-evolution-validation-contract.md and contract tests.

S53 — Minimum sufficient solution guardrail compatibility review

• Repository / layer: AletheIA readiness and planning governance
• Covers: P24-R3 minimum sufficient solution guardrail as a compatibility extension to readiness gates.
• Dependencies: S50 delivered intake and existing Readiness Gates / Work Slice Spec Bundle guidance.
• Status: delivered by AletheIA PR #371
• Non-goals: new schemas, new skills, automatic classification, policy engines, collectors, dashboards, Runtime 2.0 implementation, S18 activation or Adaptive Skills mutation.
• Acceptance: record a manual readiness gate and template fields requiring evidence before new complexity enters a Work Slice plan.
• Evidence: AletheIA PR #371 / 7554700; docs/contracts/readiness-gates-spec.md, starter-pack/templates/work-slice-readiness-review-template.md and contract tests.

S54 — Observatory interface refinement review

• Repository / layer: AletheIA Mission Control read-only UI
• Covers: P1/P3/P10/P11 Observatory maturity refinement after S32, limited to clearer read-only interpretation of existing sourced signals.
• Dependencies: S32 Observatory maturity review, S49 next-source gate and delivered S50-S53 manual compatibility path.
• Status: delivered by PR #373 and PR #374.
• Non-goals: new dashboards, collectors, live importers, ranking, scoring, comparative metrics, S18 activation, Runtime 2.0 implementation, automatic inference, policy engines or new data sources.
• Acceptance: improved the existing Resource Observatory interface so source authority, neutral unavailable state and no-decision authority are more visible without changing projection data or behavior.
• Evidence: apps/mission-control/src/views/ResourceObservatoryView.tsx, ResourceSignalInspector.tsx, apps/mission-control/src/styles.css, apps/mission-control/src/features/resource-observatory/ResourceObservatory.test.tsx, PR #373 and PR #374.

S55 — Blume official docs shell spike

• Repository / layer: AletheIA documentation tooling
• Covers: Source-backed evaluation of Blume as a static, AI-ready documentation shell for the existing AletheIA Markdown corpus.
• Dependencies: S41 documentation coherence review, S43 documentation reference integrity repair, S54 closure state and Blume documentation reviewed on 2026-07-14.
• Status: delivered as a docs-tooling spike; production publishing remains gated.
• Non-goals: replacing canonical Markdown sources, publishing production docs, changing AletheIA runtime behavior, adding dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior, Adaptive Skills mutation or automatic documentation generation.
• Acceptance: prove whether Blume can build the existing docs corpus from an isolated package; record build/validation gaps before any official-docs publishing decision.
• Evidence: apps/docs, pnpm run docs:build, pnpm run docs:validate, Blume docs at https://useblume.dev/ and https://useblume.dev/docs/quickstart.

S56 — Blume validation and source-coverage triage

• Repository / layer: AletheIA documentation tooling
• Covers: Source-backed classification of Blume docs:validate failures before any official documentation publishing decision.
• Dependencies: S55 Blume docs shell spike.
• Status: delivered; publication remains gated.
• Non-goals: publishing GitHub Pages, fixing all links at once, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: record the validation failure categories and define a safe next-step sequence for publication-source mapping and later link repair.
• Evidence: docs/reports/blume-validation-triage-s56.md, pnpm run docs:validate.

S57 — Blume publication-source map

• Repository / layer: AletheIA documentation tooling
• Covers: Decision map for which repository areas belong in the official Blume documentation site, which remain GitHub source links and which stay out of publication scope.
• Dependencies: S55 Blume docs shell spike and S56 Blume validation/source-coverage triage.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, fixing all links, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: record publication tiers and link-handling rules that unblock future small link-repair batches without turning the repository into a full docs mirror.
• Evidence: docs/reports/blume-publication-source-map-s57.md.

S58 — Blume internal docs route repair

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow repair of Blume validation failures caused by links that should resolve inside the existing docs/ source.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage and S57 publication-source map.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, expanding source coverage, fixing all links, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: reduce internal docs/ route validation failures while leaving non-docs/ source coverage decisions to later S57-category slices.
• Evidence: docs/reports/blume-internal-route-repair-s58.md, pnpm run docs:validate.

S59 — Blume root/meta and policy source-link handling

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow handling of Blume validation failures caused by repository-root, repository-meta and policy references that are canonical source files outside the docs/ Blume source tree.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map and S58 internal docs route repair.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, expanding source coverage, fixing examples/starter-pack/packs/assets, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: remove the SYSTEM_STATE.md, _meta/MIGRATION.md and policies/feature-readiness-gates.md validation category by treating them as explicit GitHub source links.
• Evidence: docs/reports/blume-root-meta-policy-source-links-s59.md, pnpm run docs:validate; PR #385.

S60 — Blume Visual Operations example source-link handling

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow handling of Blume validation failures caused by docs/ links to repository-level examples/visual-operations/ source files.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair and S59 root/meta/policy source-link handling.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, moving examples into docs/, fixing starter-pack/packs/non-Visual-Operations examples/assets, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: remove the examples/visual-operations validation category by treating those example artifacts as explicit GitHub source links.
• Evidence: docs/reports/blume-visual-ops-example-source-links-s60.md, pnpm run docs:validate; PR #387.

S61 — Blume starter-pack source-link handling

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow handling of Blume validation failures caused by docs/ links to repository-level starter-pack/ templates, guides and harness shims.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling and S60 Visual Operations example source-link handling.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, moving starter-pack content into docs/, fixing packs/non-starter examples/assets, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: remove the starter-pack/ validation category by treating those operational artifacts as explicit GitHub source links.
• Evidence: docs/reports/blume-starter-pack-source-links-s61.md, pnpm run docs:validate; PR #389.

S62 — Blume pack source-link handling

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow handling of Blume validation failures caused by docs/ links to repository-level packs/operating-overlay/ delivery pack material.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling, S60 Visual Operations example source-link handling and S61 starter-pack source-link handling.
• Status: delivered; publishing remains gated.
• Non-goals: publishing GitHub Pages, changing Blume config, moving packs into docs/, fixing non-Visual-Operations examples/assets, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: remove the packs/operating-overlay validation category by treating pack material as explicit GitHub source links.
• Evidence: docs/reports/blume-pack-source-links-s62.md, pnpm run docs:validate; PR #391.

S63 — Blume non-Visual-Operations example source-link handling

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow handling of remaining Blume broken-link validation failures caused by docs/ links to repository-level non-Visual-Operations examples/ artifacts.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling, S60 Visual Operations example source-link handling, S61 starter-pack source-link handling and S62 pack source-link handling.
• Status: delivered; publishing remains gated pending asset-check review.
• Non-goals: publishing GitHub Pages, changing Blume config, moving examples into docs/, fixing asset strategy, broad documentation rewrite, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: remove the remaining examples/ broken-link validation category by treating example artifacts as explicit GitHub source links.
• Evidence: docs/reports/blume-example-source-links-s63.md, pnpm run docs:validate; PR #393.

S64 — Blume asset-check warning readiness review

• Repository / layer: AletheIA documentation tooling
• Covers: Narrow review of the remaining BLUME_ASSETS_UNCHECKED warning after S63 cleared broken-link validation errors.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling, S60 Visual Operations example source-link handling, S61 starter-pack source-link handling, S62 pack source-link handling and S63 non-Visual-Operations example source-link handling.
• Status: delivered; official publishing remains gated pending a deliberate asset-publication strategy or documented acceptance of the warning.
• Non-goals: publishing GitHub Pages, changing Blume config, copying or relocating assets, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: document the meaning of the remaining Blume asset warning, identify the tracked local asset sources involved, and keep the publication gate explicit.
• Evidence: docs/reports/blume-asset-warning-readiness-s64.md, pnpm run docs:validate; PR #395.

S65 — Blume asset-check warning acceptance decision

• Repository / layer: AletheIA documentation tooling
• Covers: Decision on whether the remaining BLUME_ASSETS_UNCHECKED warning blocks publication-readiness planning after S64.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling, S60 Visual Operations example source-link handling, S61 starter-pack source-link handling, S62 pack source-link handling, S63 non-Visual-Operations example source-link handling and S64 asset-check warning readiness review.
• Status: delivered; official publishing remains gated by a separate publication-readiness decision.
• Non-goals: publishing GitHub Pages, changing Blume config, creating apps/docs/public/, copying or relocating assets, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: document the asset warning as accepted for the current spike if build evidence shows docs-local images are emitted as optimized static assets.
• Evidence: docs/reports/blume-asset-warning-decision-s65.md, pnpm run docs:validate, pnpm run docs:build; PR #397.

S66 — Blume publication-readiness decision

• Repository / layer: AletheIA documentation tooling
• Covers: Decision on hosting target, expected base URL, base path and publication gates for a future Blume documentation deployment.
• Dependencies: S55 Blume docs shell spike, S56 validation/source-coverage triage, S57 publication-source map, S58 internal docs route repair, S59 root/meta/policy source-link handling, S60 Visual Operations example source-link handling, S61 starter-pack source-link handling, S62 pack source-link handling, S63 non-Visual-Operations example source-link handling, S64 asset-check warning readiness review and S65 asset-check warning acceptance decision.
• Status: delivered; no deployment or workflow changes are included.
• Non-goals: publishing GitHub Pages, changing Blume config, changing repository Pages settings, creating deployment workflow, automatic documentation generation, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: document the intended future hosting target and the required gates for a separate publishing implementation slice.
• Evidence: docs/reports/blume-publication-readiness-decision-s66.md, pnpm run docs:validate, pnpm run docs:build; PR #399.

S67 — Blume GitHub Pages readiness implementation

• Repository / layer: AletheIA documentation tooling
• Covers: Minimal GitHub Pages readiness implementation after S66, including Pages base configuration, manual-only deployment workflow and smoke-test checklist.
• Dependencies: S55 Blume docs shell spike through S66 publication-readiness decision.
• Status: delivered; no automatic deployment is included.
• Non-goals: automatic GitHub Pages publishing on merge, repository Pages settings change outside Git, custom domain, generated documentation pipeline, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: local docs validation/build pass with the Pages base configuration, the workflow is workflow_dispatch only, and the smoke-test checklist is documented.
• Evidence: docs/reports/blume-pages-readiness-s67.md, pnpm run docs:validate, pnpm run docs:build; PR #401.

S68 — Blume GitHub Pages deploy smoke test

• Repository / layer: AletheIA documentation tooling
• Covers: Evidence from the first explicit manual GitHub Pages deployment and S67 smoke test.
• Dependencies: S67 Blume GitHub Pages readiness implementation and explicit human approval to run the manual workflow.
• Status: delivered; automatic publishing on merge remains disabled.
• Non-goals: automatic GitHub Pages publishing on merge, custom domain, generated documentation pipeline, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: successful manual workflow run and smoke-test evidence for home page, representative route, assets, llms.txt, robots.txt, sitemap and search index.
• Evidence: docs/reports/blume-pages-deploy-smoke-test-s68.md, GitHub Actions run 29390178825, https://nevitonsantana.github.io/AletheIA/; PR #403.

S69 — Blume docs output audit and postprocess guard

• Repository / layer: AletheIA documentation tooling
• Covers: Published Blume artifact hygiene for duplicate first headings, internal Markdown links and raw Markdown files.
• Dependencies: S68 public Pages smoke test and review of the generated GitHub Pages surface.
• Status: delivered; automatic publishing on merge remains disabled.
• Non-goals: source Markdown rewrite, automatic GitHub Pages publishing on merge, custom domain, generated documentation pipeline, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: pnpm run docs:build runs postprocess/audit and the resulting apps/docs/dist has zero raw Markdown artifacts, zero internal .md/.mdx links and zero duplicate first headings.
• Evidence: docs/reports/blume-output-audit-s69.md, scripts/postprocess-blume-docs.mjs, scripts/audit-blume-docs-output.mjs; PR #404.

S70 — Blume Pages post-guard smoke test

• Repository / layer: AletheIA documentation tooling
• Covers: Manual GitHub Pages publication after S69 and public smoke-test evidence for duplicate-heading/raw-Markdown hygiene.
• Dependencies: S69 Blume docs output audit and postprocess guard, S69 closure via PR #405 and explicit human approval to run the manual workflow.
• Status: delivered; automatic publishing on merge remains disabled.
• Non-goals: automatic GitHub Pages publishing on merge, custom domain, generated documentation pipeline, dashboards, collectors, scoring, ranking, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: manual workflow completes successfully and sampled public routes show rendered HTML pages without duplicate first headings or internal raw Markdown links.
• Evidence: docs/reports/blume-pages-post-guard-smoke-s70.md, GitHub Actions run 29391565432, https://nevitonsantana.github.io/AletheIA/; PR #406.

S71 — Docs language and ADR navigation audit

• Repository / layer: AletheIA documentation tooling/content governance
• Covers: Small public-docs coherence review for mixed English/Portuguese content and explicit Blume navigation metadata for ADR grouping/order.
• Dependencies: S70 public post-guard smoke test and human review of the published docs surface.
• Status: delivered; language normalization remains a separate follow-up decision.
• Non-goals: bulk translation, rewriting historical evidence, automatic language classification, automatic publishing on merge, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: Blume build/validate accepts meta.ts navigation metadata; ADR group is explicitly labeled and ordered; mixed-language candidates are recorded for a future reviewed translation/normalization slice.
• Evidence: docs/reports/docs-language-and-adr-navigation-audit-s71.md, docs/meta.ts, docs/adr/meta.ts; PR #407.

S72 — Public docs English language policy

• Repository / layer: AletheIA documentation content governance
• Covers: Official language decision for published public documentation before any translation/normalization work.
• Dependencies: S71 docs language and ADR navigation audit plus explicit user decision that English is the official public documentation language.
• Status: delivered; translation work remains a separate follow-up slice.
• Non-goals: bulk translation, automatic language classification, rewriting historical evidence, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: public-doc language policy records English as the official language and defines boundaries for preserving historical evidence.
• Evidence: docs/reference/public-doc-language-policy.md; PR #409.

S73 — Core concept English normalization

• Repository / layer: AletheIA public documentation content
• Covers: First small English-normalization slice for public concept pages identified by S71.
• Dependencies: S72 public docs English language policy.
• Status: delivered; further mixed-language pages remain follow-up work.
• Non-goals: bulk translation, rewriting historical evidence, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/concepts/governance.md and docs/concepts/quality.md are normalized to English while preserving meaning and existing public-doc scope.
• Evidence: docs/concepts/governance.md, docs/concepts/quality.md; PR #411.

S74 — Launch Kit English normalization

• Repository / layer: AletheIA public documentation content
• Covers: English-only normalization of the public Launch Kit after S72 established English as the canonical docs language.
• Dependencies: S72 public docs English language policy and S73 first concept normalization.
• Status: delivered; further mixed-language pages remain follow-up work.
• Non-goals: bulk translation, rewriting historical evidence, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/reference/launch-kit.md is English-only and no longer carries duplicate Portuguese announcement copy in the canonical public docs surface.
• Evidence: docs/reference/launch-kit.md; PR #413.

S75 — Architecture flow language normalization

• Repository / layer: AletheIA public documentation content
• Covers: Residual Portuguese flow label in the public architecture concept page.
• Dependencies: S72 public docs English language policy and S74 Launch Kit normalization.
• Status: delivered; further mixed-language pages remain follow-up work.
• Non-goals: bulk translation, rewriting historical evidence, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/concepts/architecture.md uses English for the core contract flow label.
• Evidence: docs/concepts/architecture.md; PR #415.

S76 — ADR-007 language normalization

• Repository / layer: AletheIA public ADR documentation
• Covers: Residual Portuguese terms and one typo in ADR-007’s APM packaging decision narrative.
• Dependencies: S72 public docs English language policy and S75 architecture flow language normalization.
• Status: delivered; broader ADR language review remains follow-up work.
• Non-goals: rewriting ADR decisions, changing APM strategy, bulk translation, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/adr/ADR-007-apm-packaging-strategy.md uses English for the decision narrative terms touched by this slice and fixes the visible nativ ely typo.
• Evidence: docs/adr/ADR-007-apm-packaging-strategy.md; PR #417.

S77 — KGL prep language normalization

• Repository / layer: AletheIA public roadmap documentation
• Covers: Residual Portuguese phase label in the Knowledge Governance Layer technical implementation prep roadmap.
• Dependencies: S72 public docs English language policy and S76 ADR-007 language normalization.
• Status: delivered; visual/Kanban PT-BR labels remain a separate follow-up because they may be communication labels rather than accidental language drift.
• Non-goals: changing KGL implementation scope, rewriting roadmap history, translating visual labels, bulk translation, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/roadmaps/knowledge-governance-implementation-prep.md uses English for the Phase 5 implementation-prep label.
• Evidence: docs/roadmaps/knowledge-governance-implementation-prep.md; PR #419.

S78 — Public language residual classification

• Repository / layer: AletheIA public documentation governance
• Covers: Classification of remaining Portuguese-language surfaces after S77.
• Dependencies: S72 public docs English language policy and S77 KGL prep language normalization.
• Status: delivered; no further automatic translation is admissible from this scan.
• Non-goals: translating historical learning records, rewriting pilot/local guides, changing visual/Kanban labels, bulk translation, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: a review report identifies remaining language categories and records the decision boundary for future editorial action.
• Evidence: docs/reports/public-language-residual-classification-s78.md; PR #421.

S79 — Open Dependabot trust-boundary review

• Repository / layer: AletheIA dependency maintenance governance
• Covers: S28/S29 advisory review of open Dependabot PRs #422, #423 and #424 before any branch update or merge.
• Dependencies: S35 Domain Governance Usage Evidence Capture, S36/S45 Dependabot trust-boundary usage records and S78 public language residual classification.
• Status: delivered; branch updates and merge decisions remain separate follow-up work.
• Non-goals: updating Dependabot branches, merging dependency PRs, validating upstream release notes as authoritative, creating scanners, policy engines, runtime enforcement, dashboards, schemas, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: S35-style evidence records the applicable packs, reviewed PRs, observed trust boundaries, current CI posture and the branch-update-before-merge decision boundary.
• Evidence: examples/domain-governance-usage/s79-open-dependabot-trust-boundary-review.json; PR #426.

S80 — Public language residual resolution

• Repository / layer: AletheIA public documentation governance
• Covers: Explicit editorial resolution for the remaining language residual categories classified by S78.
• Dependencies: S72 public docs English language policy, S78 public language residual classification and S79 dependency-maintenance closure.
• Status: delivered; no automatic translation or broad historical rewrite was introduced.
• Non-goals: bulk translation, changing public docs language policy, rewriting governance decisions, automatic language classification, publishing automation, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: docs/reference/learnings.md, docs/guides/pilot-crisis-monitor-overlay-handoff.md and docs/roadmaps/evolution-plan.md have explicit English-first treatment for the S78 residuals, while localized labels are preserved only as labeled historical examples.
• Evidence: docs/reports/public-language-residual-resolution-s80.md; PR #428.

S81 — Post-S80 backlog gate review

• Repository / layer: AletheIA backlog governance
• Covers: Safe next-step posture after S80 resolved known public-language residuals.
• Dependencies: S80 public language residual resolution and PR #429 S80 closure.
• Status: delivered; no implementation authority was introduced.
• Non-goals: publishing automation, bulk language work, S18 metrics, Runtime 2.0 implementation, dashboards, collectors, scanners, policy engines, automatic classification or Adaptive Skills mutation.
• Acceptance: repository status records the post-S80 posture, keeps Active: none as the default after closure, and identifies only source-backed future triggers.
• Evidence: docs/reports/post-s80-backlog-gate-s81.md; PR #430.

S82 — Blume Pages post-S81 publication smoke test

• Repository / layer: AletheIA public documentation publication evidence
• Covers: Manual GitHub Pages publication after S80/S81 and the PR #432 Blume/twoslash TypeScript compatibility fix.
• Dependencies: S80 public language residual resolution, S81 post-S80 backlog gate review and PR #432 docs publishing compatibility fix.
• Status: delivered; automatic publication remains blocked.
• Non-goals: automatic publishing, generated documentation authority, dashboards, collectors, scanners, policy engines, S18 metrics, Runtime 2.0 behavior or Adaptive Skills mutation.
• Acceptance: publication run evidence and representative HTTP 200 smoke tests are recorded without changing the manual publishing boundary.
• Evidence: docs/reports/blume-pages-post-s81-smoke-s82.md; PR #433.

S83 — Portable Continuity Reference Intake

• Repository / layer: AletheIA reference intake and backlog planning.
• Covers: P25 portable continuity, runtime-boundary metadata, Restart Package reuse and the source pack’s broader loop/graph/topology proposals as external planning data.
• Dependencies: S2 reference intake protocol, S82 closure, current Restart Package/handoff/finalization contracts and ADR-016.
• Status: delivered by AletheIA PR #454; no implementation authority was introduced.
• Non-goals: a Portable Governed Work State, canonical schema, composer, runtime adapter, template change, pilot execution, loop/graph/topology expansion, Runtime 2.0, memory writeback, automatic routing, observatory changes or Adaptive Skills mutation.
• Acceptance: register checksum and provenance; map requirements to existing contracts; record adopted, absorbed, deferred and rejected elements; preserve Restart Package as the only continuity artifact; authorize at most S84 compatibility review.
• Evidence: examples/reference-intake/s83-portable-continuity-intake.yaml; tests/examples/test-s83-portable-continuity-intake.test.ts; PR #454.

S84 — Restart Package portability compatibility review

• Repository / layer: AletheIA continuity compatibility review.
• Covers: whether runtime origin/destination, freshness, invalidation, sensitivity, redaction and unsupported semantics fit existing Restart Package fields or justify one optional block.
• Dependencies: accepted S83 intake and its proceed_to_compatibility_review gate.
• Status: delivered by AletheIA PR #456; no template, schema, composer, adapter, pilot or Adaptive Skills change was introduced.
• Verdicts: absorbed_by_existing_contract, refine, defer or reject.
• Non-goals: editing templates, creating schemas/composers/adapters, running a pilot or changing Adaptive Skills.
• Acceptance: complete a requirement-to-contract matrix with canonical_artifact_count: 0; S85 remains blocked until S84 records a bounded residual gap.
• Evidence: docs/reports/s84-restart-package-portability-compatibility-review.md; PR #456.

S85 — Codex-to-Codex Restart Package portability pilot

• Repository / layer: AletheIA project-specific continuity evidence.
• Covers: read-only baseline/candidate post-resume checks from the same real Work Slice snapshot in two fresh Codex sessions.
• Dependencies: S84 verdict confirming one optional portability-metadata gap.
• Status: delivered by AletheIA PR #458; read-only same-runtime evidence only, with no mutation by receiving sessions.
• Evidence: docs/pilots/s85-codex-to-codex-restart-pilot.md; PR #458.
• Non-goals: file mutation by receiving sessions, transcript replay, schema/template canonization, automation, runtime configuration or comparative claims.
• Acceptance: freeze the source commit and authoritative state; compare the current Restart Package with a candidate portability block; retain complete session outputs and independent review against authoritative sources.

S86 — Same-runtime portability value decision

• Repository / layer: AletheIA continuity evidence decision.
• Covers: whether the S85 candidate adds value over the current Restart Package.
• Dependencies: complete real S85 baseline/candidate evidence.
• Status: delivered by AletheIA PR #460 with insufficient_evidence; no template/schema/adapter authority was introduced.
• Evidence: docs/reports/s86-same-runtime-portability-value-decision.md; PR #460.
• Verdicts: accept_optional_profile, refine_and_repeat, insufficient_evidence or reject.
• Acceptance: zero critical constraint or sensitivity loss, correct next action, improvement in at least two restart indicators, and no more than ten minutes of candidate preparation overhead.

S87 — Optional Restart Package portability block

• Repository / layer: AletheIA handoff/finalization templates.
• Dependencies: S86 accept_optional_profile.
• Status: deferred.
• Non-goals: changing the minimum handoff schema, creating a new artifact, composer or runtime adapter.
• Acceptance: add one optional block only where a real runtime or trust boundary exists; absent values remain unavailable.

S88 — Bidirectional cross-runtime portability pilot

• Repository / layer: AletheIA project-specific continuity evidence.
• Dependencies: accepted S87 optional block.
• Status: deferred.
• Routes: Codex to Claude Code and Claude Code to Codex, read-only at the post-resume checkpoint.
• Non-goals: Adaptive Skills changes, instruction enforcement claims, automatic runtime configuration or execution automation.
• Acceptance: preserve constraints, validation, sensitivity and unsupported semantics in both routes with complete evidence and independent review.

S89 — Restart Package portability adoption decision

• Repository / layer: AletheIA continuity governance.
• Dependencies: S88 cross-runtime evidence.
• Status: deferred.
• Outcomes: retain as an optional Restart Package extension, absorb required fields into existing guidance, refine through another bounded pilot or remove the experiment.
• Non-goals: authorizing a composer, schema, runtime adapter, Runtime 2.0, automatic routing, memory writeback or Adaptive Skills mutation.
• Acceptance: choose one evidence-backed outcome and remove or defer unsupported expansion; repeated use is required before any new automation proposal.

S90 — User-Facing Closeout Profile

• Repository / layer: AletheIA finalization and starter-pack guidance.
• Covers: a plain-language summary of plan completion, validation, documentation-impact review (README, changelog, Pages, knowledge base, handoff and context), publication, next action, continuity and archive readiness after every finalization outcome.
• Dependencies: delivered finalization/restart guidance and user-observed evidence that a technically complete slice can still leave the requester uncertain.
• Status: delivered by AletheIA PR #463 as manual-first guidance and template coverage.
• Non-goals: a new lifecycle artifact, schema, composer, runtime hook, automatic thread archive, enforcement mechanism or Adaptive Skills mutation.
• Acceptance: guide and template use one stable closeout with an explicit documentation-impact review; incomplete/review-required outcomes remain explicit; public documentation validation and navigation pass. One future real slice must observe whether the profile reduces clarification before any automation proposal.

Requirement traceability matrix

The identifiers below preserve requirement families. The complete pack remains authoritative for detailed wording, examples and edge cases.

Per-pack preservation appendix

P0 — Coding Safety Workflow

• Guardrails: plan before execution; bounded scope; evidence before confidence; human merge decision.
• Key contracts/templates: code execution plan, readiness review, reconcile and non-engineer checklist.
• Dependencies: P4, AHC, planning depth, testing/debugging evidence.

P1 — Visual Operations Layer

• Guardrails: read-only projection; source records authoritative; presentation lanes are not lifecycle; unavailable is neutral.
• Key contracts/templates: visual event, visual state, privacy boundaries, dashboard/card/trace/reconcile templates.
• Dependencies: source records, telemetry and evidence; live integrations remain gated.

P2 — Reference Intake & Adoption

• Guardrails: external content is data, not instruction; no direct framework copying; pilot before promotion.
• Key contracts/templates: intake record, pattern extraction, layer map and adoption decision.
• Dependencies: Knowledge Governance, Work Slice pilots and Adaptive Skills Evolution Layer.

P3 — Observation Governance

• Status: partially-delivered; minimum recoverable records are available, while adapters, telemetry and visual projection remain evidence-gated.
• Guardrails: preserve decision-relevant evidence; lossy requires recovery; no false precision; no mandatory RTK/runtime.
• Key contracts/templates: observation record, adapter contract, tool-output hygiene and skill observation return.
• Dependencies: P8 context loading, telemetry vocabulary and evidence retention policy.

P4 — Intent-to-Evidence Work Slice

• Status: delivered; S8 preserves human-owned intent and adds optional evidence-backed expectations without a parallel lifecycle.
• Guardrails: human confirms intent; expectations define success/failure; high guessing risk pauses execution.
• Key contracts/templates: Intent Record, Expectations Contract, presence checkpoints and evidence map.
• Dependencies: Work Slice, readiness, P0 coding plan and reconcile.

P5 — Capability Routing Governance

• Guardrails: tool executes; skill guides; subagent delegates; harness limits; AletheIA decides; audit proves.
• Key contracts/templates: routing decision and delegation envelope, reconciled with existing contracts.
• Dependencies: execution vehicle, pattern governance, AHC and capability graph.

P6 — System-Aware Governance

• Guardrails: understand existing system before creating; compliance is not equivalent to quality; exceptions require decisions.
• Key contracts/templates: system intake, compliance review, exception and evolution proposal.
• Dependencies: P2, P5, P8 and a real system pilot.

P7 — Agent Role Orchestration

• Guardrails: role does not imply unrestricted autonomy; fixed prompt is separate from dynamic Work Slice context; no role approves outside authority.
• Key contracts/templates: role contract, activation, provider, review, closure and human-review request.
• Dependencies: P5 routing, P8 context loading and P3 evidence return.

P8 — Context Surface Governance

• Guardrails: load only when relevant; record owner/scope/freshness/budget; avoid prompt, skill and provider leakage.
• Key contracts/templates: surface registry, routing policy, smell catalog, loading policy and closure context list.
• Dependencies: current context-rot and Knowledge Governance contracts.

P9 — Governed Loop Observation

• Status: delivered; the observe milestone and first bounded debugging pilot are both accepted.
• Guardrails: observe before pilot; no UI authority; no inferred compatibility; objective gate, budget, limits and human boundary are mandatory.
• Key contracts/templates: patternContext, debugging pilot record, inspector states and loop-run evidence.
• Dependencies: PR #250, ADR-015, Adaptive Skills compatibility declarations, P3 observation return.

P10 — Work Observatory

• Status: partially-delivered; five real derived records exist, and PR #313 adds the comparison-readiness review that keeps calibrated and comparative capabilities deferred because the records are not one stable comparison group.
• Guardrails: work is not output volume; Resource Observatory remains the raw-resource source; unknown review state stays unknown; do not score people; no fake ROI.
• Canonical position: separate logical sublayer inside Observability & Intelligence, initially sharing existing visual surfaces rather than creating a new product area.
• Key future interfaces: derived work_record, comparison validity, manual acceptance/rework and human-effort provenance.
• Dependencies: P3 observations, resource telemetry, execution/audit records, S7 real evidence and privacy review.
• Deferred: work units, value scoring, comparative dashboards, model/skill rankings and Decision Intelligence.

P11 — Cognitive Documentation & Continuity

• Status: delivered baseline; SYSTEM_STATE, continuity, proportional closure and S17 explainable first use are delivered. Observatory health projection remains evidence-gated.
• Guardrails: understanding is explicit but cannot replace technical proof; plain language cannot hide risk; no closure self-certification; no compaction that loses intent, evidence or open risk.
• Absorbed foundations: comprehension debt, Durable Decisions, Context Surface Registry, Handoff Record, Slice Finalization and Restart Package.
• Key future interfaces: small per-repository SYSTEM_STATE, Restart Package continuity extension, proportional documentation-coherence report and Explainable Change Brief.
• Dependencies: S8 Intent-to-Evidence, current finalization/restart guidance, documentation indexes and human-review criteria.
• Deferred: automatic doc generation, mandatory ceremony for every Lite slice and documentation-health projection without records.

P12 — Work Observatory duplicate/continuation marker

• Status: absorbed; treated as the same Work Observatory source family as P10.
• Guardrails: no second observatory lifecycle, no duplicate metric authority and no comparative claims without denominator evidence.
• Destination: S18 remains the only comparative work-metrics review gate.

P13 — Human Expertise & Evidence-Based Learning

• Status: delivered; start with the smallest Domain Expertise Brief and Evidence-Based Learning Record.
• Guardrails: AletheIA amplifies human expertise; it does not replace human decision, approve high-risk work or let skills self-evolve.
• Dependencies: S20 skill governance and S21 validation hardening.
• Evidence: AletheIA PR #289 / 0800cd0; human expertise and evidence-based learning contract, starter-pack templates and synthetic example.
• Deferred: capability fitness dashboards, rankings and learning automation.

P14 — Governed Loop Engineering Addendum

• Status: delivered; reconciled with existing execution-pattern and loop contracts.
• Guardrails: loops must be verifiable, bounded, observable when recurring and reviewable; no objective gate means no autonomous loop.
• Dependencies: S10 routing clarity, S21 validation hardening and existing maker-checker/objective-gate policies.
• Evidence: AletheIA PR #283 / 0b0531a; S23 Loop Readiness Gate contract, review checklist, template, synthetic fixture and guardrail tests.
• Deferred: schedulers, policy engines and automatic loop execution.

P15 — Design System Intelligence / Pulso Lab

• Status: delivered; use v0.2 as authoritative for the S24 Pulso pilot slice and as the delivered evidence for the original S11 Design System pilot.
• Guardrails: Pulso is a lab; statistics are signals, not promotion decisions; who proposes does not approve.
• Dependencies: S10 routing, S20 skill governance and current Pulso source registry when the pilot starts.
• Evidence: Adaptive Skills PR #81 / a7c5dc2; AletheIA PR #287 / e180f0a; AletheIA Design System Intelligence pilot contract, starter-pack template and synthetic Pulso review; Adaptive Skills design-system-intelligence skill, capability route and example review.
• Deferred: weekly scanners, automated remediation and cross-DS generalization until a manual pilot proves value.

P16 — Independent Validation Hardening

• Status: delivered; prioritize its guardrails before broader automation.
• Guardrails: independent critic is not proof, not hidden chain-of-thought sharing and not mandatory ceremony for low-risk work.
• Dependencies: existing Intent-to-Evidence, maker-checker, AHC/AHGE and closure governance.
• Evidence: AletheIA PR #281 / fb9f900; S21 contract, blind-review checklist, starter-pack template, synthetic review case and contract tests.
• Deferred: benchmark framework, autonomous merge/deploy and universal critic requirement.

P17 — Lean Skill Doctrine

• Status: delivered; S20 delivered the AletheIA and Adaptive Skills governance halves, while S22 added the bounded implementation skill and observation compatibility.
• Guardrails: prevent skill sprawl, overlap, context bloat and mini-framework skills; block only narrow authority/safety violations.
• Dependencies: S19 source registry; S22 depends on S20.
• Evidence: Adaptive Skills PR #79 / 95408c4; AletheIA PR #285 / 9bc12f9; quality-gate record, projection status and observation-compatibility fixture.
• Deferred: skill-catalog-review skill, automatic skill routing and existing skill rewrites.

P18 — AletheIA Runtime 2.0

• Status: delivered for boundary review; ADR-016 keeps it as a strategic north-star and boundary-review source.
• Guardrails: Runtime 2.0 cannot silently move AletheIA from macro-governance into execution substrate ownership.
• Destination: S26 preserves the north-star, rejects immediate implementation and requires a later explicit boundary decision before any runtime track.
• Deferred: runtime kernel, SDK, CLI, event bus, plugin interface, provider adapters and capability runtime.

P24 — Learning Distillation & Context Hygiene Addendum v0.7

• Status: delivered for intake only; v0.7 is the current package and supersedes v0.1-v0.6 variants of the same addendum.
• Guardrails: preserve validated learning, not raw conversation history; route learning only after evidence, classification and destination review; prevent new references from silently hijacking active work; treat complexity as cost before adding new surfaces.
• Candidate destinations: existing Restart Package, SYSTEM_STATE index, finalization guidance, AletheIA-to-Adaptive-Skills evolution signals and future source-backed backlog slices. S51 reviewed the Restart Package compatibility destination; S52 reviewed the Adaptive Skills evolution-signal bridge; S53 reviewed the minimum-sufficient-solution planning guardrail.
• Dependencies: S2 reference intake, S49 next-source gate, existing continuity/finalization contracts and Adaptive Skills PR-governed evolution.
• Deferred: learning automation, automatic memory writeback, canonical schema adoption, new skills, Adaptive Skills mutations, dashboards, collectors and runtime enforcement.

P25 — Portable Continuity / Restart Package Portability v0.1

• Status: delivered for S83 intake only; the package has no implementation authority.
• Guardrails: Restart Package remains the sole continuity artifact; portability is a metadata hypothesis, not a new lifecycle; missing state stays unavailable; instruction files are context, not enforcement.
• Candidate destination: one optional Restart Package portability block, only if S84 proves an unabsorbed gap and S85-S86 prove value over the current baseline.
• Dependencies: S2 reference intake, S82 closure, existing handoff/finalization/runtime contracts and ADR-016.
• Deferred: S86 until separate explicit activation after S85; all template changes and cross-runtime work until their preceding gate.
• Rejected: new canonical state/schema/composer, Runtime 2.0, automatic routing, memory writeback, broad loop/graph/topology expansion, dashboards, collectors, comparative metrics and Adaptive Skills mutation.

M1 — Ecosystem Territory Map

• Posture: versioned north-star explanation, not a contract, ADR, implementation inventory or lifecycle.
• Interpretation: agent roles and outcomes are directional; dashboard values are illustrative; numbered regions are compositional.
• Authority: current contracts, decisions, records and evidence override the image.

Validation contract for every slice

Every implementation slice must report:

1. repository and owning layer;
2. pack requirement families covered;
3. dependencies and baseline ref;
4. explicit non-goals;
5. acceptance criteria and evidence produced;
6. risks, deferred requirements and any status changes;
7. applicable repository checks.

Minimum checks:

• AletheIA governance validation and git diff --check;
• schema, JSON and YAML validation when structured artifacts change;
• Mission Control typecheck, tests, build, snapshots and inspector accessibility regression when UI changes;
• Adaptive Skills skill, evolution, projection and capability validators when that repository changes;
• link/reference validation for all changed documentation.

Reopening and change control

Before starting a pack or supplemental slice:

1. verify the source archive checksum or register the newer revision;
2. reopen the complete pack, including guardrails, templates, examples and assessment;
3. compare it with the current repositories instead of trusting this snapshot alone;
4. update requirement statuses only with evidence;
5. create a small Work Slice and preserve unresolved requirements in this backlog.

This prevents the backlog from becoming a lossy summary while keeping day-to-day planning readable.