continuous-learning-v2

フックを介してセッションを観察し、信頼度スコアリング付きのアトミックなインスティンクトを作成し、スキル/コマンド/エージェントに進化させるインスティンクトベースの学習システム。

Quality

48%

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

Fix and improve this skill with Tessl

tessl review fix ./docs/ja-JP/skills/continuous-learning-v2/SKILL.md

Quality

Content

50%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 is well-structured and offers genuinely executable setup snippets, but it is overlong for a SKILL.md, lacks validation checkpoints around the evolve step, and critically references core scripts (observe.sh, Python CLI) that are not bundled. Splitting detail into reference files and shipping the missing scripts would lift every dimension.

Suggestions

Bundle the referenced observe.sh, hooks/hooks.json, and Python CLI under scripts/ (or references/) so the runtime the body describes actually exists.

Add an explicit validation/feedback checkpoint to the evolve workflow (e.g. review generated instincts, confirm confidence before promoting to evolved/), since clustering into skills/commands/agents is a batch operation.

Move the full config reference, command catalog, and architecture rationale into separate reference files and keep SKILL.md as a lean overview with one-level-deep links; fix the broken ```json fence in the plugin-install section.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with concrete config/commands, but sections like the v1/v2 comparison table and the 'why hooks not skills' rationale explain design history Claude does not need to act, and could be trimmed.	3 / 5
Actionability	Inline guidance is concrete (settings.json hooks, mkdir, config.json, instinct YAML, slash commands), but the system's core runtime — observe.sh, hooks/hooks.json, and the Python CLI — is referenced yet absent from the bundle, leaving a key execution gap.	3 / 5
Workflow Clarity	The quick start is sequenced (enable hooks → init dirs → use commands), but no validation or feedback checkpoints exist for the batch-like evolve/clustering operation, capping workflow clarity at 3; a malformed ```json fence in the plugin-install section also muddies the steps.	3 / 5
Progressive Disclosure	Section headers give reasonable structure, but the ~250-line SKILL.md is monolithic with config reference, command docs, and architecture all inlined, and it references bundle files (scripts/references) that do not exist.	3 / 5
	Total	12 / 20 Passed

Description

46%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 conveys a clear and distinctive 'what' but is written in the skill's own jargon and omits any 'when to use' guidance. Adding natural trigger phrases and an explicit Use-when clause would materially raise completeness and trigger-term quality.

Suggestions

Append an explicit 'Use when...' trigger clause (e.g. when the user wants Claude to learn their recurring patterns or build reusable behaviors from session observations).

Replace internal jargon ('instinct', 'atomic', 'confidence scoring') in the first clause with natural user-facing terms ('learned behaviors', 'patterns', 'reliability scoring') so users would actually say them.

Lead with the concrete user-visible outcomes before the architectural mechanism to improve specificity.

Dimension	Reasoning	Score
Specificity	Names the domain ('instinct-based learning system') and a few concrete actions — 'observes sessions via hooks', 'creates atomic instincts', 'evolves them into skills/commands/agents' — but these are architectural/system verbs rather than tangible user-facing operations, leaving coverage non-comprehensive.	3 / 5
Completeness	It clearly states what the system does, but there is no 'Use when...' clause or equivalent trigger guidance, so completeness is capped at 3 per the rubric guidelines.	3 / 5
Trigger Term Quality	The phrasing leans on internal jargon ('instinct', 'confidence scoring', 'atomic') with only generic terms like 'learning system' and 'observe sessions'; it lacks the natural phrases a user would actually say when they need this skill.	2 / 5
Distinctiveness Conflict Risk	The hook-driven, confidence-scored instinct pipeline is a fairly distinct niche with low overlap risk against typical skills, though the broad 'learning system' framing leaves minor overlap risk.	4 / 5
	Total	12 / 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
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: affaan-m/ECC
Path: docs/ja-JP/skills/continuous-learning-v2/SKILL.md
Commit: e04ea0b

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.