continuous-learning

自动从Claude Code会话中提取可重复使用的模式，并将其保存为学习到的技能以供将来使用。

2.25x

Quality

51%

Does it follow best practices?

Impact

90%

2.25x

Average score across 3 eval scenarios

Securityby

High

Do not use without reviewing

Fix and improve this skill with Tessl

tessl review fix ./docs/zh-CN/skills/continuous-learning/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 with concrete config and hook setup, but it inlines a sizable v2 research/comparison section that belongs in a separate file and leaves the core extraction logic abstract. Adding explicit validation checkpoints and moving the v2 material out would raise quality.

Suggestions

Move the Homunculus v2 comparison and 'potential v2 enhancements' section into docs/continuous-learning-v2-spec.md, leaving a one-line pointer in SKILL.md to improve conciseness and progressive disclosure.

Add an explicit validation/review checkpoint to the workflow (e.g. 'Before saving a learned skill, confirm it is non-trivial and matches a detectable pattern category; skip if it falls under ignore_patterns').

Surface the core extraction logic—either inline the key steps of evaluate-session.sh or give a concrete summary of how patterns are detected and saved—so the skill is executable without reading an external script.

Dimension	Reasoning	Score
Conciseness	Mostly efficient (config, hook setup, pattern table are tight), but the ~30-line Homunculus v2 comparison and '研究：2025年1月' research section is inlined bloat that does not help execute the current skill and could be trimmed.	3 / 5
Actionability	Provides concrete, copy-paste-ready config and hook-setup JSON, but the core extraction logic is described only at a high level ("识别可提取的模式", "保存有用的模式") and delegated to an unseen evaluate-session.sh, leaving key execution detail incomplete.	3 / 5
Workflow Clarity	The three steps (session evaluation → pattern detection → skill extraction) are sequenced, but there are no explicit validation checkpoints; the auto_approve review is an implicit config flag rather than a stated workflow gate.	3 / 5
Progressive Disclosure	Section structure is clear and well-organized, but the v2 comparison/research content that clearly belongs in the referenced docs/continuous-learning-v2-spec.md is inlined, and no bundle files exist to verify referenced paths.	3 / 5
	Total	12 / 20 Passed

Description

53%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 states what the skill does but omits any explicit trigger/when guidance, which caps completeness and limits activation reliability. It is reasonably specific and distinct, yet could be strengthened with natural trigger phrases.

Suggestions

Add an explicit 'Use when...' clause with concrete trigger phrases, e.g. 'Use when the user wants to learn from or review Claude Code sessions, extract reusable patterns, or manage saved learned skills.'

Broaden trigger-term coverage with natural synonyms users would say (e.g. 'learn from sessions', 'save what I learned', 'review session patterns').

Mention the pattern categories (error resolution, corrections, workarounds, debugging) in the description to lift specificity toward 4-5.

Dimension	Reasoning	Score
Specificity	Names two concrete actions ("提取可重复使用的模式" / extract reusable patterns, and "保存为学习到的技能" / save as learned skills) but is not comprehensive—no mention of what kinds of patterns or how analysis happens.	3 / 5
Completeness	Has a clear 'what' (extract patterns from sessions and save as learned skills) but no 'when'/'Use when...' clause, which the guideline explicitly caps at 3.	3 / 5
Trigger Term Quality	Contains some relevant keywords ("Claude Code会话", "可重复使用的模式", "学习到的技能") but misses common natural variations and synonyms a user would actually say, and lacks a 'Use when' trigger phrase.	3 / 5
Distinctiveness Conflict Risk	Defines a clear niche (session pattern extraction → learned skills) that is mostly distinct with only minor overlap risk against related learning/meta skills; not a 5 because the missing explicit trigger weakens reliable activation.	4 / 5
	Total	13 / 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/zh-CN/skills/continuous-learning/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.