skill-creator

Create or update GoClaw agent skills with eval-driven iteration. Use for new skills, skill scripts, references, benchmark optimization, description optimization, eval testing, extending agent capabilities.

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./skills/skill-creator/SKILL.md

Quality

Content

65%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 SKILL.md body is a lean, actionable overview with concrete commands and a strong reference structure, but it loses points for duplicated publish content, ungated validation in its batch workflows, and several dangling bundle references (agents/, eval-viewer/) that resolve to no real files.

Suggestions

Consolidate the publish_skill guidance into one place (either the Creation Workflow step or the 'Publishing to System' section) to remove duplication and tighten conciseness.

Add an explicit validation gate to the eval/batch workflow — e.g. 'Only proceed to Optimize when assertions pass; if a run fails, fix the skill and re-run' — to lift workflow clarity.

Resolve the dangling references: either add the missing agents/grader.md, agents/comparator.md, agents/analyzer.md and eval-viewer/ files, or repoint those citations to existing bundles (scripts/generate_report.py, assets/eval_review.html).

Dimension	Reasoning	Score
Conciseness	The body is dense and imperative with no basic-concept padding, but it duplicates publish_skill across the Creation Workflow step 8 and the 'Publishing to System' section, and the Quick Reference / Scripts tables overlap inline citations, so it could be tightened past the level-3 bar.	2 / 3
Actionability	It gives copy-paste-ready commands ("scripts/init_skill.py <name> --path <dir>", "publish_skill(path: \"~/.goclaw/skills-store/<name>\")") and a full script-to-purpose table, matching the fully-executable anchor.	3 / 3
Workflow Clarity	The 10-step Creation Workflow and 7-step Eval workflow are clearly sequenced, but validation checkpoints are implicit (a 'Test & Evaluate' step with no 'only proceed when tests pass' gating) and the batch eval operation lacks an explicit fix-and-retry feedback loop, capping it at 2.	2 / 3
Progressive Disclosure	The overview is well-organized with clearly signaled one-level-deep references, but body-cited paths to a missing agents/ directory (grader.md, comparator.md, analyzer.md) and a missing eval-viewer/ directory are dangling against the actual bundle, breaking easy navigation.	2 / 3
	Total	9 / 12 Passed

Description

75%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 cleanly answers both what and when with an explicit 'Use for' trigger clause and a distinctive skill-creation niche, but its capability list and trigger terms lean on internal categories (benchmark/description optimization, eval testing) rather than the concrete actions and natural phrasings a user would say.

Suggestions

Replace category-style use cases ('benchmark optimization', 'description optimization', 'eval testing') with concrete actions a user would say, e.g. 'optimize a skill's trigger description', 'run benchmark evals', 'improve a skill's accuracy score'.

Add common natural phrasings users actually say (e.g. 'make a new skill', 'write a skill script', 'package a skill for distribution') to broaden trigger coverage.

Lead the action clause with specific verbs beyond 'Create or update' — e.g. 'write, evaluate, optimize, and publish' — to lift specificity from level 2 to 3.

Dimension	Reasoning	Score
Specificity	"Create or update GoClaw agent skills with eval-driven iteration" names the domain and core actions, but the listed use cases ("benchmark optimization, description optimization, eval testing") are categories rather than multiple specific concrete actions, so it falls short of level 3.	2 / 3
Completeness	It states what ("Create or update GoClaw agent skills with eval-driven iteration") and provides an explicit trigger clause ("Use for new skills, skill scripts, references, benchmark optimization..."), satisfying both what and when with explicit guidance.	3 / 3
Trigger Term Quality	"new skills", "skill scripts", and "references" are natural user terms, but "benchmark optimization", "description optimization", and "eval testing" lean toward internal jargon and miss common phrasings, matching the 'some relevant keywords but missing common variations' anchor.	2 / 3
Distinctiveness Conflict Risk	"GoClaw agent skills" scoped to skill creation is a clear niche with distinct triggers unlikely to overlap with unrelated skills, matching the level-3 anchor.	3 / 3
	Total	10 / 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: nextlevelbuilder/goclaw
Path: skills/skill-creator/SKILL.md
Commit: e71d6a5

Reviewed: 10 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.