Visual Operations Layer

Goal

Explain how AletheIA can make governed AI-assisted work visible without creating a second authority, runtime, or project-management system.

The Visual Operations Layer is a read-only projection over records that already describe a Work Slice: framing, decisions, execution, validation, handoff, restart, telemetry, audit, and learning.

Its interface name is AletheIA Mission Control.

The layer may normalize, aggregate, filter, and present existing signals. It must not authorize an action, change a readiness outcome, close a slice, promote a learning, or become the canonical store for any of those decisions.

AletheIA contracts and project records
  Work Slice · readiness · decisions · evidence · audit · handoff · learning
        ↓ references / normalizes
Visual Operations projection
  board lanes · cards · trace · alerts · telemetry summaries
        ↓ renders
AletheIA Mission Control
  read-only operational legibility

Layer boundaries

Layer	Responsibility	Visual Operations relationship
AletheIA	Frames the Work Slice, governs readiness, risk, decisions, continuity, and learning	Source authority
Adaptive Skills	Declares reusable capabilities, expected evidence, verification, and handoff signals	Optional capability signal
Runtime / harness	Executes within an approved envelope and emits observations or audit records	Optional execution signal
Visual Operations	Projects source records into human-readable views	Read-only consumer

A skill activation can be shown, but it cannot decide the gate. A runtime event can be shown, but it cannot prove success without evidence. A board lane can summarize the current posture, but it cannot replace the source record that established that posture.

Existing authorities

The first projection reuses these existing surfaces:

Work Slice Pattern for the bounded operational unit and derived states;
Readiness Gates Spec for continue, tighten, review, handoff, escalate, and stop outcomes;
Slice Telemetry Model and Context / Resource Telemetry Spec for provider-agnostic resource signals;
Agent Harness Governance Extension for trace events emitted by a harness;
Agent Action Audit Record and Execution Audit Record for reviewable action and execution evidence;
decision, execution, handoff, restart, and learning artifacts attached to the Work Slice.

Visual Operations documents define how to read these sources together. They do not redefine their fields or precedence.

Projection principles

Source references are mandatory. Every derived state, alert, and evidence item points to the record that supports it through source_refs.
Unknown remains unknown. Missing data is rendered as unknown or unavailable; the projection never invents precision.
Telemetry carries provenance. Token and cost values declare whether they are reported, estimated, or unavailable.
Metadata first. Sensitive content stays in its governed source; the projection carries a reference, classification, hash, or authorized summary.
Human review is first-class. Pending review, the reviewer role, and the open question remain visible until the authoritative source resolves them.
Ceremony is proportional. A small slice may emit only a card and a few events; richer views are optional when the risk or evidence burden justifies them.

What this layer is not

not a runtime, scheduler, orchestrator, or policy engine;
not a replacement for GitHub Issues, Projects, Linear, Jira, or another coordination board;
not a mandatory state machine for every task;
not a store for full prompts, secrets, personal data, or restricted source content;
not a productivity ranking based on tokens, cost, speed, or skill usage;
not a claim that a dashboard is stronger evidence than the records it references.

First delivery boundary

The docs-first delivery consists of projection contracts, starter templates, and synthetic examples. It does not generate a dashboard, ingest runtime logs, connect to an external service, or implement a user interface. Those require separate Work Slices after the vocabulary has been reviewed in use.