moai-harness-learner

Harness learning subsystem coordinator. Produces Tier 4 auto-update proposal payloads consumed by the orchestrator (which surfaces them via AskUserQuestion) and orchestrates Apply/Rollback flows. Triggers when harness learning proposals are pending or learning lifecycle management is needed.

Quality

66%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Fix and improve this skill with Tessl

tessl review fix ./.claude/skills/moai-harness-learner/SKILL.md

Quality

Content

53%Weight 40%Scale 1-5

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

The body gives concrete, mostly-executable commands and a well-structured workflow, but it is padded with policy-history annotations and repeated contract citations, and its destructive Apply/Rollback steps lack explicit validation feedback loops.

Suggestions

Remove the HTML policy annotation block and the V3R3/V3R4 policy-history parentheticals; keep only the behavioral contract, cited once.

Consolidate the repeated "Canonical contract: askuser-protocol.md § Orchestrator-Subagent Boundary" references into a single mention in Quick Reference.

Add an explicit validation checkpoint to Step 4 (e.g. re-run `moai harness status` to confirm the proposal was applied and no frozen-path or canary violation occurred) before declaring success, and a verification step after rollback.

Dimension	Reasoning	Score
Conciseness	Noticeable padding: a large HTML policy annotation (the V3R4/V3R3 contract paragraph), policy-history meta-commentary ("the harness-learning policy, superseded by the harness foundation policy..."), and the "Canonical contract" citation restated in Quick Reference, Step 3, and the Safety Architecture all earn no actionable value.	2 / 5
Actionability	Concrete, mostly-executable commands (`moai harness status/apply/rollback/disable --project-root`) and a documented JSON payload schema are provided, but "Run `moai harness apply --execute` (if the CLI supports it) or call the harness API directly" is a hedge rather than a fully executable instruction.	4 / 5
Workflow Clarity	The status→apply→payload→approve/reject sequence is listed, but the workflow involves destructive/batch operations (applying auto-updates, deleting proposal files, rollback) with no explicit validate→fix→retry checkpoints in the steps, capping this dimension at 3 per the destructive-operation rule.	3 / 5
Progressive Disclosure	Content is organized into clear sections (Quick Reference, Implementation Guide steps, Works Well With, Safety Architecture Reference) with signaled references; no bundle files exist so nesting is not a concern, though the repeated contract citations could be consolidated into one.	4 / 5
	Total	13 / 20 Passed

Description

80%Weight 40%Scale 1-5

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 clearly and explicitly answers both what the skill does and when to trigger it, with a distinct niche and multiple concrete actions. Its main weakness is trigger-term naturalness — the keywords are subsystem jargon rather than user-facing language.

Dimension	Reasoning	Score
Specificity	"Produces Tier 4 auto-update proposal payloads", "consumed by the orchestrator (which surfaces them via AskUserQuestion)" and "orchestrates Apply/Rollback flows" name several concrete actions, though it does not enumerate the individual CLI subcommands it coordinates, leaving a minor coverage gap versus the 5 anchor.	4 / 5
Completeness	It explicitly states what the skill does (produces payloads, orchestrates Apply/Rollback) and when to use it ("Triggers when harness learning proposals are pending or learning lifecycle management is needed"), satisfying both halves with a concrete trigger phrase.	5 / 5
Trigger Term Quality	The trigger clause names "harness learning proposals are pending" and "learning lifecycle management", but these are internal subsystem jargon rather than the natural phrases a user would spontaneously say, and common synonyms/variations are missing.	3 / 5
Distinctiveness Conflict Risk	"Harness learning subsystem coordinator" plus Tier 4 auto-update / AskUserQuestion / Apply-Rollback carve a clear, narrow niche unlikely to be confused with other skills.	5 / 5
	Total	17 / 20 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning

	Total	15 / 16 Passed

Repository: modu-ai/moai-adk
Path: .claude/skills/moai-harness-learner/SKILL.md
Commit: 4100d87

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.