skill-builder

Reviews and improves SKILL.md packages by fixing audit findings, triggers, examples, evals, token budget, release proof, safety verdicts, comparator/baseline choices, and bounded code-lens hardening. Use when the user says improve a skill, fix a skill file, review SKILL.md, raise Tessl score, reduce context cost, add skill evals, or prepare a plugin skill for release.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with a clear, validated workflow and real one-level references, but it is verbose: identical release-evidence rules are restated across multiple sections and heavy procedural detail that could live in references is kept inline. Tightening the repetition and moving the proof-ladder/release-evidence detail into a reference would lift both conciseness and progressive disclosure.

Suggestions

De-duplicate the release-evidence rules: state the 20-scenario, 300-run/20-reserve, scenario-drift, and handoff-readiness requirements once (e.g., in Validation) and reference them elsewhere instead of restating them in Workflow, Execution Boundaries, and Validation.

Move the full SDK proof-ladder ordering and per-gate pass criteria into a references file (e.g., references/proof-ladder.md), keeping only the next-action summary inline in SKILL.md to improve progressive disclosure.

Collapse the First-Principles Gate and Philosophy sections into a shorter rationale; several lines ('Rejected copied assumption', 'Rejected alternatives') restate constraints already covered by the Workflow and Constraints sections.

Dimension	Reasoning	Score
Conciseness	Almost every line is operational rather than concept-teaching, but the same release-evidence rules are repeated across Workflow, Validation, and Execution Boundaries (e.g., the 20-scenario rule and the 300-run/20-reserve rule each appear three times), which is verbosity the rubric penalizes.	2 / 3
Actionability	Provides copy-paste-ready commands ('./bin/ask skills external-review <target> --audit-level compat --json --robot', './bin/ask sdk start <skill-path> --json --robot'), concrete pass criteria ('status == "success"', '>= 90', '>= max(0.90, baseline)'), and a diff example — fully executable guidance.	3 / 3
Workflow Clarity	A numbered 1–6 workflow with explicit validation checkpoints ('run the focused gate; record baseline', 'rerun the same gate', 'On green, run final gates'), a flat-loop feedback mechanism, and a fail-fast rule — matching the clear-sequence-with-explicit-validation anchor.	3 / 3
Progressive Disclosure	References are clearly signaled and one-level deep ('See [repair examples](./references/repair-examples.md)', a References section, and a scripts pointer) and the referenced files exist, but a large amount of dense procedural detail (the full SDK proof ladder and all release-evidence rules) sits inline in the entrypoint rather than being offloaded.	2 / 3
	Total	10 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: concrete capability list, explicit 'Use when' triggers, third-person voice, and a distinct niche. Minor risk is that several triggers ('Tessl score', 'code-lens') are domain jargon, but they are framed as natural user phrasing.

Dimension	Reasoning	Score
Specificity	Lists many concrete repair actions — 'fixing audit findings, triggers, examples, evals, token budget, release proof, safety verdicts, comparator/baseline choices, and bounded code-lens hardening' — matching the multiple-specific-actions anchor.	3 / 3
Completeness	Explicitly answers both what ('Reviews and improves SKILL.md packages by fixing...') and when (a 'Use when...' clause with explicit triggers), so it clears the completeness-3 bar rather than capping at 2.	3 / 3
Trigger Term Quality	The 'Use when the user says improve a skill, fix a skill file, review SKILL.md, raise Tessl score, reduce context cost, add skill evals, or prepare a plugin skill for release' clause gives good coverage of phrases a user would naturally say.	3 / 3
Distinctiveness Conflict Risk	The niche is narrow (skill repair and release-proofing through the Tessl SDK proof ladder) with distinct triggers, making it unlikely to fire for unrelated skills; voice is third person.	3 / 3
	Total	12 / 12 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_field	'metadata' should map string keys to string values	Warning
referenced_paths_exist	Referenced path issues: 3 missing	Warning

	Total	14 / 16 Passed

Repository: jscraik/Agent-Skills
Commit: 5a6027f

Reviewed: 5 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.