learn

Update the project's long-term memory after a merge to main. Reads the merged diff and the current memory, then writes Expert shards, discovered invariants, candidate lints, and AGENTS.md pointers — all on a reviewable learn/<sha> PR. Use post-merge (the harness invokes it automatically) or with --rebuild to regenerate memory from scratch. Triggers - learn, expert-update, update memory, update expert, post-merge memory, self-improve (project)

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

77%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.

A well-structured orchestration skill with a clear sequenced workflow, real validation/feedback loops, and clean one-level progressive disclosure into verified reference files. Its main weakness is verbosity in the philosophy section, including some repeated phrasing.

Suggestions

Tighten the P1–P8 philosophy: collapse the duplicated 'lint is a rule the agent cannot ship past' phrasing and trim em-dashed asides.

Make the abstract Step 4 actions concrete (e.g., name the exact shard files to diff and the command used to run a drafted lint) to close the remaining actionability gap.

Dimension	Reasoning	Score
Conciseness	Mostly efficient but the P1–P8 philosophy block, repeated asides ('the highest-value memory, because a lint is a rule the agent cannot ship past' appears twice), and em-dashed parentheticals add padding that could be tightened; not verbose enough for a 2, not lean enough for a 4.	3 / 5
Actionability	Provides concrete commands and paths ('claude -p "/learn --since <sha> --sha <sha>"', 'scripts/check-agents-md.sh', 'git ls-remote origin learn/<sha>', '.claude/skills/expert/references/*.md') with minor gaps where steps stay abstract ('apply diffs to the relevant shard(s)').	4 / 5
Workflow Clarity	Steps 0–6 are explicitly sequenced with validation checkpoints — Step 4 requires the drafted lint to pass against current main ('if it fails on current main it's wrong'), Step 5 re-validates, and the 'Hard nevers' section acts as a checklist; a clear feedback loop for this batch/memory operation.	5 / 5
Progressive Disclosure	Clear overview in SKILL.md pointing to five one-level-deep references (all verified to exist: routing-rules, expert-shards, agents-md-guidance, invariant-discovery, consensus) plus the check-agents-md.sh script, each signaled with a one-line description; navigation is easy and content is appropriately split.	5 / 5
	Total	17 / 20 Passed

Description

88%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.

A specific, well-structured description that answers both what and when with an explicit trigger list. It could improve by adding natural-sounding synonyms and de-jargoning a few trigger terms.

Suggestions

Add more natural trigger synonyms (e.g., 'remember this project', 'update project docs') so users phrase it more ways.

Soften jargony triggers like 'expert-update' and 'self-improve (project)' toward phrasings a user would actually say.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions with comprehensive coverage — 'Reads the merged diff and the current memory, then writes Expert shards, discovered invariants, candidate lints, and AGENTS.md pointers' — naming four distinct write outputs plus '--rebuild to regenerate memory from scratch'.	5 / 5
Completeness	Clearly answers 'what' (update long-term memory, reads diff, writes shards/invariants/lints/AGENTS.md pointers) and 'when' via an explicit 'Use post-merge (the harness invokes it automatically) or with --rebuild' clause plus concrete trigger phrases.	5 / 5
Trigger Term Quality	Explicit 'Triggers - learn, expert-update, update memory, update expert, post-merge memory, self-improve (project)' gives good keyword coverage, but several terms are project-jargon and there are no synonyms or file extensions; not a 5.	4 / 5
Distinctiveness Conflict Risk	Pinned to a clear niche (post-merge memory reconciliation on a learn/<sha> PR), but the bare 'learn' and 'update memory' triggers are broad enough to risk minor overlap with related memory skills; mostly distinct rather than minimal-conflict.	4 / 5
	Total	18 / 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
referenced_paths_exist	Referenced path issues: 5 missing	Warning

	Total	15 / 16 Passed

Repository: tdg-ninja/context-specs-factory-ai
Path: skills/harness/learn/SKILL.md
Commit: 3a7a725

Reviewed: about 19 hours 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.