create-control-manifest

Flat must-do/never-do rules sheet per system and layer, extracted from Accepted ADRs. ADRs explain why; this is actionable.

Quality

72%

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 ./.claude/skills/create-control-manifest/SKILL.md

Quality

Content

81%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 content is a strong, highly actionable workflow document: exact commands, prompts, validation gates, and a complete output template leave little ambiguity. Its weaknesses are moderate verbosity in the config-table sections and a missing when-trigger in the description that the body only partially compensates for.

Dimension	Reasoning	Score
Conciseness	The body is dense and procedural, but there are noticeable tightenings available: the naming/performance tables repeat "from naming.* in project.yaml, else technical-preferences.md" in both the Value and Source columns for every row, and the "Manifest Version" paragraph explains that two fields hold the same date twice. This fits "mostly efficient but includes some unnecessary explanation or could be tightened" rather than the 4 anchor's "minor instances".	3 / 5
Actionability	Everything is executable: exact Grep patterns with flags and -A values, literal user-facing message strings, enumerated AskUserQuestion prompts and options, and a complete copy-paste output template with concrete file paths. Not 4 because there are no gaps — the common cases are fully specified.	5 / 5
Workflow Clarity	The phased sequence (1–6 plus the 4b gate) includes explicit validation checkpoints with feedback loops: the denominator/status-resolution table with hard stops, user approval before any write, the TD-MANIFEST director gate with APPROVE/CONCERNS/REJECT/NOT ASSESSED verdicts and fix-and-re-run paths, and regeneration/overwrite handling. This matches the anchor with explicit validation steps and error-recovery loops.	5 / 5
Progressive Disclosure	The body is well organized into clearly signaled phases, and external doc references (.claude/docs/automation-modes.md, director-gates/td-manifest.md, config-resolution.md) are one level deep and consistently named. Not 5 because the ~100-line output template is fully inline with no separate reference file for it or the layer-classification guidance; not 3 because what is inline is the artifact format itself, which legitimately belongs in the body.	4 / 5
	Total	17 / 20 Passed

Description

53%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 concrete about what the skill produces and reasonably distinct, but it never names the artifact ("control manifest") and gives no guidance on when to invoke the skill. It answers "what" well and "when" not at all.

Suggestions

Add an explicit when-clause, e.g. "Use when a control manifest is needed or when ADRs are newly accepted or revised."

Include the artifact's natural name and trigger phrases ("control manifest", "rules sheet", "regenerate manifest") so users and Claude can match on what they would actually say.

Mention the concrete actions the skill performs (extract rules per layer, classify, review via TD-MANIFEST gate, write docs/architecture/control-manifest.md) to raise specificity.

Dimension	Reasoning	Score
Specificity	The description names the domain and a concrete deliverable ("Flat must-do/never-do rules sheet per system and layer, extracted from Accepted ADRs") but covers only a single artifact rather than the several specific actions of a 4 or 5 — no mention of regenerating, reviewing, or classifying by layer.	3 / 5
Completeness	The "what" is clearly stated, but there is no "Use when…" clause or equivalent trigger guidance; the rubric explicitly caps completeness at 3 for a missing when-clause.	3 / 5
Trigger Term Quality	It includes relevant keywords ("must-do/never-do", "rules", "ADR", "layer") but misses the natural trigger users would actually say — notably the artifact's own name "control manifest" — as well as common phrasings like "rules sheet" or "regenerate rules".	3 / 5
Distinctiveness Conflict Risk	"Extracted from Accepted ADRs" carves a clear niche that is unlikely to trigger for unrelated skills; the only minor overlap risk is with sibling ADR/architecture-pipeline skills, matching the "mostly distinct" anchor rather than the fully distinct 5.	4 / 5
	Total	13 / 20 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
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: Donchitos/Claude-Code-Game-Studios
Path: .claude/skills/create-control-manifest/SKILL.md
Commit: b21fa0f

Reviewed: 3 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.