Skip to content
Adaptive Skills Docs
Esc
↑↓navigate↵open⌘Jpreview
On this page

Skills catalog

Browse all published Adaptive Skills by category, trigger, and expected outcome.

Adaptive Skills currently publishes 38 portable skills. Use this page when you already know the kind of help you need. If the task is still unclear, start with Choose the right skill.

Each skill appears once in this catalog. Its skills/<name>/SKILL.md file remains the canonical instruction source.

Skill Catalog: 11 Categories

Skill Catalog: 11 Categories — 38 portable skills organized by category, trigger, and expected outcome.

Engineering

Skill Use it when Expected outcome
workflow Starting non-trivial work or returning from a handoff Explicit scope, proof, and next step
feature-planning Turning a feature request into staged delivery Small, testable delivery plan
testing Choosing proof before closing meaningful work Risk-appropriate validation strategy
debugging Investigating a reproducible failure Isolated cause, fix, and recurrence guard
api-design Defining a contract between components Clear inputs, outputs, errors, and compatibility
refactoring Improving structure without changing behavior Safer, clearer structure with preserved behavior
lean-implementation Implementing a confirmed bounded plan Smallest safe change with evidence
architecture-review Reviewing a consequential structural decision Explicit alternatives, trade-offs, and recommendation
code-style Improving consistency and readability Applied rules without style dogma
communication Explaining a change, failure, option, or handoff Audience-appropriate, evidence-backed message
domain-language-alignment Product, domain, docs, and code use different terms Shared vocabulary before implementation

Design

Skill Use it when Expected outcome
ux-strategy The interaction direction is still open Compared directions and a reasoned position
ux-provocation A dominant UX hypothesis needs pressure testing Exposed assumptions and stronger alternatives
heuristic-audit An existing interface has usability risks Prioritized failures and defensible fixes
ux-writing Product language is unclear or unsafe Clear, actionable interface language
design-system-intelligence Reviewing an artifact against declared design-system sources Source-backed conformance and exception evidence
experience-design A broad experience request needs selective composition A bounded path through design building blocks

Documentation

Skill Use it when Expected outcome
documentation Creating, restructuring, or auditing onboarding, manuals, guides, and technical documentation Self-service reader journeys, source-backed content, controlled procedural clarity, and validated recovery paths

Planning

Skill Use it when Expected outcome
intent-clarification Outcome, success, or constraints are ambiguous Human-owned intent clarified before planning
premortem A consequential plan could fail in costly ways Failure causes, warning signals, and prevention

Efficiency

Skill Use it when Expected outcome
task-chunking Work is oversized or likely to sprawl Reviewable slices with stop conditions
checkpoint-review A long task needs a continue-or-stop decision Deliberate checkpoint and next-round decision
handoff-summary Work is crossing a person, agent, or session boundary Verified context for the next operator
work-continuity-efficiency A broad continuity need requires selective composition A bounded path through continuity building blocks

Cross-functional

Skill Use it when Expected outcome
triad-check A hard-to-reverse decision crosses product, design, and engineering Three-perspective decision review

Product

Skill Use it when Expected outcome
product-management A product request spans framing, evidence, method selection, outcomes, or handoffs Bounded product decision path with explicit modules, uncertainty, verification, and human boundaries
feature-value-governance Deciding whether a feature deserves investment Auditable value and overreach judgment
opportunity-tree-alignment Backlog bets lack clear outcome links Outcome-to-opportunity-to-lever alignment

Business

Skill Use it when Expected outcome
business-design Framing product or business strategy from mixed signals Structured facts, hypotheses, and implications
revenue-lever-mapping A value claim has no measurable mechanism Primary lever, metric, proxy, and risks

Governance

Skill Use it when Expected outcome
feature-complexity-audit Permanent feature cost may change a build decision Complexity scorecard and reduction recommendation
sunset-decision An existing feature may no longer justify its cost Keep, limit, refactor, deprecate, or remove decision
knowledge-source-evaluation A document may become governed knowledge Source maturity and registration decision
knowledge-conflict-resolution Decision-relevant sources disagree Precedence-based resolution or escalation
restricted-context-check Sensitive or untrusted knowledge may enter task context Leakage, injection, permission, and contamination check
knowledge-governance A broad source-governance decision needs selective composition A bounded path through governance building blocks

Quality

Skill Use it when Expected outcome
qa-review A change needs a cross-layer quality gate Consistency and operational-risk review

Metrics

Skill Use it when Expected outcome
observability-review Metrics, events, alerts, or diagnostics need review Decision-linked observable signals

What is outside this catalog

  • Domain-pack skills are validation cases and are not part of the 35-skill portable library.
  • Capability metadata is an advisory discovery layer, not an additional skill inventory.
  • Evolution records describe how skills improve; they are not executable skills.

Next steps

Was this page helpful?