CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

sources

Find, capture, score and index clinical sources for this knowledge base. Use whenever the user asks to search for or add new clinical information, adds content that makes a clinical claim, uploads a document, asks what a source is worth or where it came from, or types "/sources". Runs the WA-first locality ladder, records every source pulled including rejections, and proves the register is complete before the work is called done.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

88%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.

A tightly written, highly actionable instruction-only skill: concrete commands, exact paths, and explicit validation loops for every workflow, with no wasted tokens. The main gaps are the absent example register record format and inline scoring/publisher detail that belongs in the referenced protocol file.

Suggestions

Add a minimal example record (or field list) for src/data/source-acquisitions.json so a 'find' or 'add' capture can be written correctly without consulting the external protocol doc.

Move the band-score arithmetic ('76 unreviewed and 91 once signed off') and the unregistered-publisher list into docs/source-acquisition-protocol.md, keeping only the decision rule inline.

Verify the '../../../docs/source-acquisition-protocol.md' relative link resolves from the skill's installed location, or move the protocol into a references/ file inside the skill bundle.

DimensionReasoningScore

Conciseness

The ~80-line body is dense with project-specific rules — the locality ladder, register dispositions, band semantics, publisher registry consequences — and never explains anything Claude already knows; even justifications like 'Fabricated provenance is the worst failure this system can produce, because it looks exactly like real provenance' earn their tokens by calibrating judgment. Every line carries a rule or a consequence, matching the 'every token earns its place' anchor.

5 / 5

Actionability

Guidance is highly concrete: exact commands ('npm run check:source-acquisitions', 'npm run check:source-catalogue'), exact file paths ('src/data/source-acquisitions.json', 'src/lib/source-authority-registry.ts', 'tests/source-catalogue-providers.test.ts'), concrete expected scores ('76 unreviewed and 91 once signed off'), and a specified report table format. It falls short of 5 only because no example register record or its field format is shown inline — the entry schema is left to the protocol doc and the checker.

4 / 5

Workflow Clarity

Each verb has a numbered sequence with an explicit validation checkpoint and a feedback loop — 'Run `npm run check:source-acquisitions` and fix what it names' for find, checks for add, and a state-printing review — plus guidance for blocked paths (unregistered publishers: 'Report it to the owner... and stop'). This matches the 5 anchor's validate → fix → retry pattern applied to batch register operations.

5 / 5

Progressive Disclosure

Good structure: the body is a well-sectioned overview with the full protocol offloaded one level deep to a clearly signaled reference ('Full protocol: docs/source-acquisition-protocol.md') and rules located by path in source files. It falls short of 5 because no bundle files (references/, scripts/, assets/) exist, the protocol link points outside the skill bundle with a fragile '../../../' relative path, and inline detail like band-score arithmetic and the unregistered-publisher list could live in the referenced protocol.

4 / 5

Total

18

/

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.

An exemplary description: third-person, concrete verbs for every capability, and an explicit 'Use whenever' clause with a rich set of natural trigger phrases including a slash command. Both the what and when are answered clearly and nothing is padded or vague.

DimensionReasoningScore

Specificity

The description enumerates multiple concrete actions — 'Find, capture, score and index clinical sources', 'Runs the WA-first locality ladder, records every source pulled including rejections, and proves the register is complete' — with comprehensive coverage of the skill's capabilities. It exceeds the 4 anchor ('several specific actions; minor gaps') because no capability of the skill is left unnamed.

5 / 5

Completeness

It explicitly answers both what ('Find, capture, score and index clinical sources for this knowledge base') and when ('Use whenever the user asks to search for or add new clinical information...') with concrete trigger phrases, exactly matching the 5 anchor.

5 / 5

Trigger Term Quality

Triggers cover natural user phrasings including synonyms — 'asks to search for or add new clinical information', 'adds content that makes a clinical claim', 'uploads a document', 'asks what a source is worth or where it came from', 'types "/sources"' — giving comprehensive coverage of how a user would actually express the need. It is well above the 4 anchor ('good keyword coverage; a few natural terms missing').

5 / 5

Distinctiveness Conflict Risk

The niche is distinct — clinical source acquisition, WA-first locality ladder, source register, '/sources' command — so overlap risk with other skills is minimal. It is clearly above the 4 anchor ('minor overlap risk with closely related skills').

5 / 5

Total

20

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

relative_links

Relative link issues: 1 suspicious

Warning

Total

15

/

16

Passed

Repository

BigSimmo/PsychSift

Path

.claude/skills/sources/SKILL.md

Reviewed

Table of Contents

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.