continuous-learning

自动从 Claude Code 会话（Session）中提取可重用的模式（Pattern），并将其作为已学习的技能（Skill）保存以备将来使用。

Quality

50%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

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

Quality

Content

50%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.

The body is well-structured with concrete config and hook examples, but it bloats the context window with a dated comparison investigation and lacks the executable extraction logic it alludes to. Referenced scripts and file paths are not present in the bundle, undermining navigation.

Suggestions

Move the '对比笔记（调查：2025年1月）' comparison investigation out of SKILL.md into a separate reference file (or a clearly marked deprecated section) to recover conciseness.

Provide the actual evaluate-session.sh or its core extraction logic/prompt, and verify every referenced path (including the /Users/affoon/... file) is a real bundle file or remove it.

Add an explicit validation checkpoint to the workflow (e.g. confirm a detected pattern exceeds the extraction_threshold and is not in ignore_patterns before writing a learned skill).

Dimension	Reasoning	Score
Conciseness	The core sections (how it works, config, pattern types, hook config) are reasonably efficient, but the dated '对比笔记（调查：2025年1月）' comparison table and the '为什么使用 Stop 钩子' rationale add unnecessary tokens; the inline dated investigation content is not isolated in a deprecated section, so it cannot reach the lean level-3 anchor.	2 / 3
Actionability	Concrete artifacts are present (config.json, hook JSON, paths like ~/.claude/skills/learned/), but the actual pattern-extraction mechanism is abstract ('识别可从会话中提取的模式') and references evaluate-session.sh which is not provided in the bundle, leaving key execution details incomplete.	2 / 3
Workflow Clarity	The 工作原理 section lists a 3-step sequence (评估 → 检测 → 提取) but with no validation checkpoints or feedback loops for the skill-saving write operation; per the rubric, missing validation in batch/write workflows caps workflow clarity at 2.	2 / 3
Progressive Disclosure	Sections are organized, but the file references evaluate-session.sh and a local /Users/affoon/... path that do not exist as bundle files, and investigative content that should be separate is inline; this matches 'some structure but content that should be separate is inline' rather than clean one-level-deep references.	2 / 3
	Total	8 / 12 Passed

Description

50%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 clearly states what the skill does but omits any explicit trigger guidance for when Claude should use it, which caps both completeness and trigger quality. It is specific enough to be distinguishable but would benefit from natural user-facing 'Use when...' language.

Suggestions

Add an explicit trigger clause, e.g. 'Use when a Claude Code session ends and reusable patterns or learned skills should be captured for future use.'

Include natural keyword variations users might actually say (e.g. 'session patterns', 'learned skills', 'remember what worked') alongside the technical terms.

Lead with the concrete actions in third-person imperative voice and trim the parenthetical English glosses to tighten specificity.

Dimension	Reasoning	Score
Specificity	It names the domain ("Claude Code 会话") and two actions ("提取可重用的模式", "作为已学习的技能保存") but the action set is not comprehensive, matching the 'names domain and some actions' anchor rather than the multi-action level-3 example.	2 / 3
Completeness	It answers 'what' (extract patterns, save as skills) but provides no explicit 'when' trigger; per the guidelines a missing 'Use when...' clause caps completeness at 2, which is the closest anchor.	2 / 3
Trigger Term Quality	Terms like "Claude Code 会话", "可重用的模式", and "技能" are relevant but lean technical and lack common user variations; no natural 'Use when...' phrasing is present, so it fits 'some relevant keywords but missing common variations' rather than full coverage.	2 / 3
Distinctiveness Conflict Risk	The session-pattern-learning niche is somewhat specific, but without explicit triggers it could still overlap with related learning/memory skills, matching 'somewhat specific but could still overlap' rather than the clearly-distinct level-3 anchor.	2 / 3
	Total	8 / 12 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: xu-xiang/everything-claude-code-zh
Path: docs/ja-JP/skills/continuous-learning/SKILL.md
Commit: dfbf946

Reviewed: 28 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.