skill-creator

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update 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.

Quality

79%

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

Fix and improve this skill with Tessl

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

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 workflow with strong validation and feedback loops, but it is padded with conversational asides and — more seriously — references several files and directories that are absent from the actual bundle while leaving bundled scripts unreferenced.

Suggestions

Fix broken references: the body points to eval-viewer/generate_review.py and agents/grader.md, agents/comparator.md, agents/analyzer.md, none of which exist in the bundle — either add these files or update the paths to point at scripts/generate_report.py and any real agent docs.

Tighten conciseness by removing the padded conversational asides ('Cool? Cool.', the plumbers/grandparents anecdote, 'billions a year in economic value', the repeated caps apologies) that explain motivation Claude already grasps.

Document the existing bundled scripts: generate_report.py, improve_description.py, run_eval.py, quick_validate.py, and package_skill.py are present but mostly unreferenced — add one-line pointers from the relevant workflow sections so they are discoverable.

Dimension	Reasoning	Score
Conciseness	The operational content is mostly efficient, but several padded asides ("Cool? Cool.", the plumbers/grandparents anecdote, "billions a year in economic value," repeated emphasis blocks, "Sorry in advance but I'm gonna go all caps") add unnecessary tokens, keeping it above the noticeably-verbose 2 but short of the trim 4.	3 / 5
Actionability	Concrete copy-paste JSON schemas (evals.json, eval_metadata.json, timing.json, grading.json field names) and exact bash commands (python -m scripts.aggregate_benchmark, scripts.run_loop, scripts.package_skill) cover the common cases, but some referenced entry points (eval-viewer/generate_review.py, agents/*.md) do not match the actual bundle, leaving minor gaps.	4 / 5
Workflow Clarity	The draft → test → grade → aggregate → analyst → review → improve → repeat loop is explicitly sequenced with numbered steps, validation checkpoints (grading.json, analyst pass, feedback.json), and a feedback loop via --previous-workspace, matching the anchor-5 pattern of explicit validation and error-recovery loops.	5 / 5
Progressive Disclosure	Structure is reasonable — one-level-deep references, a dedicated "Reference files" index, and schemas/templates split into references/ and assets/ — but the body repeatedly points to eval-viewer/generate_review.py and agents/*.md which do not exist in the bundle, and several bundled scripts (generate_report, improve_description, run_eval, quick_validate) are never referenced, breaking navigation and discoverability.	3 / 5
	Total	15 / 20 Passed

Description

88%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: third-person voice, explicit what-and-when structure, and a comprehensive list of concrete capabilities with natural trigger phrasings. It is held back only by a few missing casual synonyms and minor overlap with general benchmarking/eval tooling.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "Create new skills, modify and improve existing skills, and measure skill performance" plus "run evals," "benchmark skill performance with variance analysis," and "optimize a skill's description" — giving comprehensive coverage rather than the 1-2 actions of a 3.	5 / 5
Completeness	Explicitly states what ("Create new skills, modify and improve existing skills, and measure skill performance") and when ("Use when users want to create a skill from scratch... optimize a skill's description for better triggering accuracy"), matching the anchor-5 pattern of concrete trigger phrases for both.	5 / 5
Trigger Term Quality	Good natural-term coverage with synonyms ("create a skill from scratch," "update or optimize an existing skill," "run evals," "benchmark," "test a skill"), but common casual variants like "make a skill" or "build a skill" are absent, so it is not fully comprehensive at 5.	4 / 5
Distinctiveness Conflict Risk	The skill-creation niche is clear with distinct triggers ("create a skill," "optimize a skill's description"), but "benchmark skill performance with variance analysis" and "run evals" overlap with general eval/benchmark tooling, leaving minor conflict risk vs the minimal-risk anchor 5.	4 / 5
	Total	18 / 20 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: netease-youdao/LobsterAI
Path: SKILLs/skill-creator/SKILL.md
Commit: 100d3f7

Reviewed: 2 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.