generate-agent-skills

Architects, generates, and validates Agent Skills. Enforces specification and best practices. Used any time an agent skill must be created or updated.

1.25x

Quality

—

Does it follow best practices?

Impact

100%

1.25x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Quality

Content

85%

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-structured, highly actionable workflow with real commands, explicit validation checkpoints, feedback loops, and clean progressive disclosure into verified reference files. Its main weakness is conciseness — heavy repeated emphasis and redundant STOP conditions add tokens without adding information.

Suggestions

Consolidate the repeated scaffolding warnings (the "NON-NEGOTIABLE" block, the Step 3 "MANDATORY" banner, and the Step 5 STOP condition all restate the same rule) into a single statement to tighten conciseness toward level 3.

Reduce decorative emphasis (🚨/⚠️/🛑 banners and all-caps directives) where the instruction is already clear, since the rubric penalizes verbosity even when accurate.

Dimension	Reasoning	Score
Conciseness	The body is efficient in places (real commands, explicit file paths) but pads heavily with repeated emphasis — "🚨 CRITICAL", "NON-NEGOTIABLE", "MANDATORY STEP - DO NOT SKIP", and multiple STOP conditions restating the same scaffolding rule. This matches level 2's "mostly efficient but includes some unnecessary explanation or could be tightened" rather than the lean level-3 anchor where every token earns its place.	2 / 3
Actionability	It provides fully executable commands (e.g., `python3 scripts/scaffold_skill.py --name <skill-name>`, `python3 scripts/validate_skill.py --path <path>`, `ls -la .github/skills/<skill-name>/`) with concrete arg placeholders and specific reference paths, matching level 3's "fully executable code/commands; copy-paste ready."	3 / 3
Workflow Clarity	The 6-step sequence includes explicit validation checkpoints (Verification Checkpoint after Step 3, Post-Validation Checklist after Step 5), STOP conditions, and a fix-then-re-validate feedback loop, matching level 3's "clear sequence with explicit validation steps; feedback loops for error recovery; checklists."	3 / 3
Progressive Disclosure	SKILL.md is an overview that pushes detail into one-level-deep, clearly-signaled reference files (BEST_PRACTICES.md, SPECIFICATION.md, TEMPLATES.md, workflows.md, output-patterns.md) — all verified to exist — with a "Knowledge Retrieval" navigation map, matching level 3's "clear overview with well-signaled one-level-deep references; easy navigation."	3 / 3
	Total	11 / 12 Passed

Description

75%

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 the skill does and when to use it with explicit trigger guidance, giving it strong completeness and distinctiveness. Its weakness is specificity and trigger-term breadth — the listed actions are abstract and the trigger vocabulary is narrow, missing common variations a user might actually say.

Suggestions

Add more concrete capabilities to the description (e.g., scaffolding skill directories, generating frontmatter/templates, validating SKILL.md structure) to lift specificity from level 2 to level 3.

Broaden trigger-term coverage with natural phrasings like "skill", "scaffold", "SKILL.md", or "skill frontmatter" so it matches how users actually phrase the request.

Dimension	Reasoning	Score
Specificity	Names three concrete actions ("Architects, generates, and validates Agent Skills") but they remain fairly abstract and omit specific capabilities the body actually performs (scaffolding, template generation, frontmatter validation). It clears level 2's "names domain and some actions" but falls short of the highly concrete multi-action examples at level 3.	2 / 3
Completeness	It answers WHAT ("Architects, generates, and validates Agent Skills. Enforces specification and best practices.") and WHEN ("Used any time an agent skill must be created or updated") with an explicit trigger clause, matching level 3's "clearly answers both what AND when."	3 / 3
Trigger Term Quality	"Used any time an agent skill must be created or updated" provides some natural trigger language, but coverage is thin — only "created or updated" with no common variations (skill, scaffold, SKILL.md, frontmatter). This matches level 2's "some relevant keywords but missing common variations" rather than the broad term coverage at level 3.	2 / 3
Distinctiveness Conflict Risk	The "Agent Skill" niche with the trigger "agent skill must be created or updated" is distinct and unlikely to fire for unrelated skills, matching level 3's "clear niche with distinct triggers; unlikely to conflict."	3 / 3
	Total	10 / 12 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
referenced_paths_exist	Referenced path issues: 7 missing	Warning

	Total	15 / 16 Passed

Repository: canonical/copilot-collections
Commit: d1b6440

Reviewed: 4 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.