Add agent-proxy-evaluate-skill-use skill and skill-fit view #70

New issue

Closed

opened 2026-08-01 05:46:21 +00:00 by coilyco-ops · 0 comments

coilyco-ops commented

2026-08-01 05:46:21 +00:00

Member

Goal

Expose the retained Ward skill-use and agent-compose selection evidence through a governed view, then add a skill that evaluates skill adoption and observed outcomes.

Product prerequisite

The raw ledger retains ward.skill observations and selected-skill claims, but the current materializer and operational views do not preserve skill names for querying. Land the query contract before advertising the skill.

Scope

Preserve skill identity and selection or use facts through materialization.
Add a governed skill_fit view grouped by skill, role, harness, and model.
Report observed usage, trajectory counts, completion, retry, fallback, evaluation, access, freshness, and source trajectories.
Add agent-proxy-evaluate-skill-use through the shared read-only query helper.
Keep unresolved disagreement and missing evidence explicit.

Acceptance

Metadata-only and access-tier boundaries remain intact.
The view never treats correlation as causation or grants authority.
Replayed evidence produces deterministic results.
The skill can answer whether a skill was selected or used and what outcomes were observed.
Features, operational-view docs, tests, skill validation, and Ward gates pass.

## Goal Expose the retained Ward skill-use and agent-compose selection evidence through a governed view, then add a skill that evaluates skill adoption and observed outcomes. ## Product prerequisite The raw ledger retains `ward.skill` observations and selected-skill claims, but the current materializer and operational views do not preserve skill names for querying. Land the query contract before advertising the skill. ## Scope * Preserve skill identity and selection or use facts through materialization. * Add a governed `skill_fit` view grouped by skill, role, harness, and model. * Report observed usage, trajectory counts, completion, retry, fallback, evaluation, access, freshness, and source trajectories. * Add `agent-proxy-evaluate-skill-use` through the shared read-only query helper. * Keep unresolved disagreement and missing evidence explicit. ## Acceptance * Metadata-only and access-tier boundaries remain intact. * The view never treats correlation as causation or grants authority. * Replayed evidence produces deterministic results. * The skill can answer whether a skill was selected or used and what outcomes were observed. * Features, operational-view docs, tests, skill validation, and Ward gates pass.

coilyco-ops added the

priority

autonomy

headless

labels

2026-08-05 10:25:50 +00:00

coilyco-ops referenced this issue from a commit

2026-08-12 11:16:34 +00:00

Preserve skill identity and add the governed skill-fit view

coilyco-ops referenced this issue from a pull request that will close it,

2026-08-12 11:16:35 +00:00

Preserve skill identity and add the governed skill-fit view #96