CtrlK
BlogDocsLog inGet started
Tessl Logo

skill-creator

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

67

Quality

81%

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

70%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 a well-sequenced, actionable guide to skill authoring and eval-driven iteration with strong workflow clarity. It is held back by conversational padding that pushes it past its own length budget and by several broken bundle references (agents/, eval-viewer/) that undermine both conciseness and navigation.

Suggestions

Fix or remove broken references: the cited `eval-viewer/generate_review.py` is actually `scripts/generate_report.py`, and the `agents/grader.md`, `agents/comparator.md`, `agents/analyzer.md` files do not exist in the bundle — either add them or repoint to where the equivalent guidance lives.

Trim conversational filler (e.g. 'Cool? Cool.', the plumbers/grandparents aside, the 'billions in economic value' line, and the all-caps viewer rant) to bring the body back under its recommended 500 lines.

Move the FastClaw runtime note and the platform-specific (Claude.ai / Cowork) sections into a short reference file, keeping SKILL.md as a lean overview with clear pointers.

DimensionReasoningScore

Conciseness

Mostly efficient core instructions (concrete commands, JSON templates, directory layouts) but padded with conversational asides — 'Cool? Cool.', the 'plumbers and grandparents' tangent, 'we are trying to create billions a year in economic value', and the all-caps rant — plus a FastClaw runtime detour. This fits the 3-anchor ('mostly efficient but includes some unnecessary explanation or could be tightened'); not the 2-anchor since the bulk is genuinely useful, not just padded prose.

3 / 5

Actionability

Provides mostly executable guidance — `python -m scripts.aggregate_benchmark`, `nohup python <skill-creator-path>/eval-viewer/generate_review.py ...`, exact JSON field names (text/passed/evidence) — but several commands point to paths that don't resolve (`eval-viewer/generate_review.py` vs the real `scripts/generate_report.py`), matching the 4-anchor ('mostly executable; minor gaps') rather than the fully-copy-paste 5-anchor.

4 / 5

Workflow Clarity

The eval→review→improve process is explicitly sequenced (Step 1 through Step 5) with checkpoints (grade assertions, aggregate into benchmark, analyst pass, launch viewer), a feedback loop ('improve → rerun → review → repeat'), and termination criteria ('Keep going until the user is happy / feedback empty / no progress'), matching the 5-anchor ('clear sequence with explicit validation steps; feedback loops; checklists').

5 / 5

Progressive Disclosure

Has real structure — a dedicated 'Reference files' index, organized bundle dirs (references/schemas.md and assets/eval_review.html both exist) — but the body runs ~527 lines (over its own 500-line guidance) and repeatedly cites files that don't exist (agents/grader.md, agents/comparator.md, agents/analyzer.md, eval-viewer/generate_review.py), so navigation breaks. This fits the 3-anchor ('some structure but could be better organized; references present but not reliably signaled') rather than the 4-anchor whose gaps are only minor.

3 / 5

Total

15

/

20

Passed

Description

92%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 strong: it comprehensively enumerates concrete capabilities and pairs them with an explicit, well-scoped 'Use when...' clause covering creation, editing, optimization, eval/benchmark, and description-tuning. Its only soft spot is a couple of internal-jargon trigger phrases that real users rarely say verbatim.

DimensionReasoningScore

Specificity

Lists multiple concrete actions with comprehensive coverage: 'Create new skills, modify and improve existing skills, and measure skill performance' plus 'create a skill from scratch, edit, or optimize an existing skill, run evals... benchmark skill performance... optimize a skill's description'. This matches the 5-anchor ('multiple specific concrete actions; comprehensive coverage') and is not the 4-anchor since no meaningful capability gaps remain.

5 / 5

Completeness

Explicitly answers both 'what' (create/modify/improve/measure skills) and 'when' ('Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals...') with concrete trigger phrases, matching the 5-anchor and exceeding the 4-anchor whose 'when' is only loosely specified.

5 / 5

Trigger Term Quality

Good natural-term coverage ('create a skill from scratch', 'edit', 'optimize an existing skill', 'run evals', 'test a skill', 'benchmark'), but phrases like 'variance analysis' and 'better triggering accuracy' lean technical/internal rather than user-spoken, and casual synonyms like 'make a skill' are absent — fits the 4-anchor ('good coverage; a few natural terms missing') rather than the comprehensive 5-anchor.

4 / 5

Distinctiveness Conflict Risk

Occupies a clear niche (skill authoring + eval tooling) with distinct triggers unlikely to fire for unrelated skills, matching the 5-anchor ('clear niche with distinct triggers; minimal conflict risk'); not the 4-anchor because overlap risk with adjacent skills is negligible.

5 / 5

Total

19

/

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.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (528 lines); consider splitting into references/ and linking

Warning

Total

15

/

16

Passed

Repository
fastclaw-ai/fastclaw
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.