continuous-learning-v2

Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

65%

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

The content is highly actionable with concrete commands and configs, but it is a long single-file document that lacks verification checkpoints in its setup workflow and references bundle files that do not exist. Splitting detail into the referenced files and adding a hook-verification step would improve both conciseness and workflow clarity.

Suggestions

Add a verification step to the quick start, e.g. run a tool call and confirm an entry appears in observations.jsonl, so the hook setup has an explicit checkpoint.

Move the conceptual rationale (hooks-vs-skills, confidence-evolution philosophy) into a separate reference file and keep SKILL.md as a lean overview, realizing the progressive-disclosure structure the references imply.

Either supply the referenced bundle files (observe.sh, hooks/hooks.json, start-observer.sh, config.json) in scripts/references or note explicitly that they live elsewhere, since the body cites paths that are absent from the bundle.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient with tables, config, and commands, but the ~260-line length includes conceptual rationale ("為何 Hooks vs Skills 用於觀察?", confidence-evolution philosophy) that could be tightened, and the large ASCII pipeline diagram adds bulk.	2 / 3
Actionability	Provides copy-paste-ready hook JSON config, concrete mkdir/touch commands, a complete config.json example, and a file-structure tree — fully executable guidance rather than abstract description.	3 / 3
Workflow Clarity	The quick start is numbered and sequenced (enable hooks → init directories → run observer), but there are no validation or verification checkpoints (e.g. confirming the hook actually fires), so checkpoints are implicit.	2 / 3
Progressive Disclosure	Sections are well organized, but content is monolithically inline in one ~260-line file and references bundle files (observe.sh, hooks/hooks.json, agents/start-observer.sh) that are not present in references/scripts/assets, so the split is not actually realized.	2 / 3
	Total	9 / 12 Passed

Description

60%

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 is specific and concrete about capabilities but omits explicit "when to use" triggers and relies on jargon-heavy phrasing that users are unlikely to voice naturally. Adding a "Use when..." clause with plain-language triggers would lift the completeness and trigger-term scores.

Suggestions

Add an explicit trigger clause, e.g. "Use when you want Claude to learn your recurring patterns, remember corrections, or build reusable skills from your sessions."

Replace jargon like "instinct-based" and "evolves them" with plain user-facing terms such as "learns from your sessions" and "turns them into reusable skills, commands, and agents".

Include natural keyword variations users would actually say (e.g. "learn my patterns", "remember my corrections", "build a skill from what I keep doing").

Dimension	Reasoning	Score
Specificity	Lists several concrete actions — "observes sessions via hooks", "creates atomic instincts with confidence scoring", and "evolves them into skills/commands/agents" — rather than vague abstractions.	3 / 3
Completeness	It clearly answers what the skill does but provides no "Use when..." clause or equivalent explicit trigger guidance, so the "when" is missing; per the rubric this caps completeness at 2.	2 / 3
Trigger Term Quality	It names a few relevant terms (sessions, hooks, learning) but leans on technical jargon like "instinct-based", "atomic instincts", and "evolves" that users would rarely say naturally, missing common variations.	2 / 3
Distinctiveness Conflict Risk	The instinct/evolution niche is fairly specific, but without natural trigger terms it could still overlap with other learning or memory skills, so it is not clearly conflict-free.	2 / 3
	Total	9 / 12 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: ysyecust/everything-claude-code
Commit: a9cbecb

Reviewed: 1 day 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.