CtrlK
BlogDocsLog inGet started
Tessl Logo

apm-spec-guardian

Use this skill to run a four-panel adversarial advisory review on any pull request that touches the OpenAPM specification artifact (docs/src/content/docs/specs/openapm-*.md), its inline / sidecar JSON Schemas (docs/src/content/docs/specs/schemas/*.schema.json), or the conformance fixture seed (tests/fixtures/spec-conformance/**). The panel fans out to four spec-ecosystem reviewers (swagger-openapi-editor, oci-distribution-editor, pkgmgr-registry-contract-editor, w3c-tag-architect), each running in its own agent thread, and a spec-editor synthesizer that produces a fold-now / defer-v0.1.1 / defer-v0.2 / reject list plus a ship decision keyed off a 1..10 shocked_meter scale. The orchestrator is the sole writer to the PR: ONE consolidated comment, no verdict labels, no merge gating. The panel is advisory -- it surfaces findings, prioritizes folds, and renders a ship recommendation that the maintainer weighs.

74

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

85%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable orchestration spec with clear wave sequencing, explicit validation checkpoints, and properly split bundle references. Its one weakness is repetition of the advisory-regime invariants across multiple sections, which inflates length without adding new guidance.

Suggestions

Consolidate the advisory-regime statements: state the 'no verdict / no gate / panel is advisory' invariant once in 'Architecture invariants' and reference it from Waves 3-4, the Output contract, and Gotchas rather than restating it verbatim each time.

Trim the Gotchas section of items that rehash invariants already covered above (e.g. blocker-veto and calibrated-severity are restated), keeping only the genuinely new pitfalls like spec-drift across count sites and per-byte ASCII enforcement.

Consider moving the full Wave 3 panelist prompt fragments into an asset (e.g. assets/panelist-prompt.md) so the body stays a lean overview and the prompt text lives one level deep with the other bundle materials.

DimensionReasoningScore

Conciseness

Mostly efficient and free of general-knowledge padding, but the advisory regime ('no verdict', 'does not gate merge', 'panel is advisory') is restated across the intro, architecture invariants, Wave 3, Wave 4, Output contract, and Gotchas, and several gotchas rehash invariants already stated above. Not 3 because the repetition could be tightened; not 1 because there is no explanatory fluff about concepts Claude already knows.

2 / 3

Actionability

Highly concrete guidance: numbered waves, exact task invocations, exact paths, the literal marker 'apm-spec-guardian: new-version', line thresholds ('< 50 lines'), and the exact label 'spec-review' to sweep via remove-labels. Instruction-only but specific and copy-ready per the code_vs_instruction_skills note.

3 / 3

Workflow Clarity

Waves 0→6 are clearly sequenced with a topology diagram and explicit validation checkpoints (schema-validate each panelist return, re-spawn max 2 attempts, re-spawn synthesizer once on schema failure, mandatory linter before render, loop budgets with exhaustion semantics). Feedback loops are present for the parallel/batch and PR-state-write operations.

3 / 3

Progressive Disclosure

Overview body points one level deep to real, verified bundle files — assets/panelist-return-schema.json, assets/synthesizer-return-schema.json, assets/linter-checklist.md, assets/comment-template.md — each clearly signaled by role. No nested 2+ level reference chains within the bundle; content is appropriately split between the orchestration overview and the asset files.

3 / 3

Total

11

/

12

Passed

Description

100%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concrete, third-person, and explicitly scopes activation to OpenAPM spec/schema/fixture PRs with a clear 'Use when' trigger. It names specific actions and distinguishes itself from the general docs-sync skill. Minor risk that internal mechanism terms (shocked_meter, panelist roster) dilute user-facing trigger language, but the trigger scope remains clear.

DimensionReasoningScore

Specificity

Lists multiple concrete actions: 'run a four-panel adversarial advisory review', 'fans out to four spec-ecosystem reviewers', 'produces a fold-now / defer-v0.1.1 / defer-v0.2 / reject list plus a ship decision keyed off a 1..10 shocked_meter scale', and 'ONE consolidated comment, no verdict labels, no merge gating'.

3 / 3

Completeness

Opens with an explicit 'Use this skill to run...' trigger and enumerates the exact in-scope paths, answering both 'what' (four-panel advisory review + synthesizer + single comment) and 'when' (PR touches spec artifact / schemas / fixtures). Not below 3 because both halves are explicit rather than implied.

3 / 3

Trigger Term Quality

Enumerates natural trigger terms a user would say — 'pull request', 'OpenAPM specification artifact', 'JSON Schemas', 'conformance fixture' — with concrete glob patterns. Not the level below because the activation scope is spelled out in user-recognizable vocabulary rather than only internal jargon.

3 / 3

Distinctiveness Conflict Risk

Narrow niche (OpenAPM spec artifact only) with an explicit exclusion — 'Edits to any OTHER documentation page MUST NOT trigger this skill' — making conflict with sibling skills unlikely.

3 / 3

Total

12

/

12

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 5 suspicious

Warning

Total

15

/

16

Passed

Repository
microsoft/apm
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.