self-awareness

Claude Science's own session database schema and SDK surface for introspection via host.query(). Load this when you need to query your own conversation history, token usage, cost accounting, execution log, or artifact metadata beyond what host.frames()/host.artifacts() provide — e.g. "how many tokens has this session used", "what was my last tool call", "list every file I've written", "where are messages stored", "what tables can I query", "inspect frames.context_data", or any time you're about to PRAGMA-probe the Claude Science metadata DB to discover its schema.

Quality

93%

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

86%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 an excellent, lean, actionable reference with executable examples and no wasted tokens, but it is monolithic: SKILL.md functions as the full detailed catalog rather than an overview that progressively discloses to separate reference files. Splitting the per-table schemas and SDK table into reference files would improve navigability.

Suggestions

Move the detailed per-table column listings (Queryable tables, Execution history, Compute and verification) into a references/schema.md file and keep SKILL.md as an overview with a one-line pointer plus the dialect/limits and worked examples.

Extract the SDK surface table into a references/sdk.md (or references/sdk.md) and link to it from a short 'SDK surface' summary section, so the core SKILL.md stays scannable.

Add a brief 'Quick start' section at the top with the single most common query (token/cost accounting) before the full schema reference, giving a clear progressive-disclosure entry point.

Dimension	Reasoning	Score
Conciseness	Dense, information-packed reference with no explanation of concepts Claude already knows; every line earns its place (table columns, dialect limits, caps, denylist) and operational asides are domain-specific guidance, not padding.	5 / 5
Actionability	Four fully executable, copy-paste SQL worked examples cover the common cases (token/cost accounting, recent cells, context depth, artifacts), plus concrete introspection commands like PRAGMA table_info and sqlite_master queries.	5 / 5
Workflow Clarity	Operational sequencing is clear — run via the repl tool, scope to project, narrow columns to avoid the cell/size caps, aggregate server-side — but there is no explicit multi-step validation/retry loop, which is acceptable for read-only queries yet keeps it below a 5.	4 / 5
Progressive Disclosure	The file is well-organized with clear section headers, but the entire detailed schema catalog and SDK surface are inlined in a ~260-line SKILL.md rather than an overview pointing to one-level-deep reference files; no bundle files exist to offload the detail.	3 / 5
	Total	17 / 20 Passed

Description

100%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 is exemplary: it names concrete capabilities, supplies natural-language trigger phrases users would actually say, and cleanly separates what the skill does from when to invoke it. It is distinct from sibling skills and avoids fluff.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete introspection actions — query conversation history, token usage, cost accounting, execution log, artifact metadata, and PRAGMA-probe schema discovery — giving comprehensive coverage rather than vague language.	5 / 5
Completeness	Explicitly states both what ('Claude Science's own session database schema and SDK surface for introspection via host.query()') and when ('Load this when you need to query...') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Embeds natural verbatim phrases a user would say ('how many tokens has this session used', 'what was my last tool call', 'list every file I've written', 'what tables can I query') alongside technical triggers, covering synonyms and variations.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (the Claude Science metadata DB) and explicitly differentiates from host.frames()/host.artifacts(), giving it distinct triggers with minimal conflict risk.	5 / 5
	Total	20 / 20 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: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/claude-science/self-awareness/SKILL.md
Commit: 99df6e4

Reviewed: 6 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.