An exemplary procedural workflow skill: fully executable commands and templates, rigorously sequenced phases with validation and error-recovery at every decision point, and sound delegation of policy to external docs. The only real cost is length — several justificatory asides push token cost above what the instructions alone require.

Dimension	Reasoning	Score
Conciseness	The bulk is lean, procedural instruction with exact commands and templates; however, several rationale passages could be trimmed — "The previous form of this step read the current GDD in full and git show-ed the full committed version..." and "an unbounded Read on a large ADR hits the 25k-token cap and, unrecovered, the only path forward is paging through the entire remainder" explain failure modes at length. Not 3: over-explanation is minor relative to the mostly efficient instruction; not 5: the meta-commentary about a "previous form of this step" and repeated why-not explanations are trimmable.	4 / 5
Actionability	Fully executable throughout: exact commands ("git diff HEAD -- design/gdd/[filename].md"), complete Grep invocations with pattern, glob, and output_mode, copy-ready report and dialog templates, and concrete thresholds (~50KB, 25k-token cap). Specific examples cover the common cases, including the ambiguity-resolution table for zero-match scans.	5 / 5
Workflow Clarity	Ten clearly numbered phases with explicit validation checkpoints at every branch: argument/file validation in step 1, empty-diff disambiguation in step 2 ("An empty diff is not 'no impact'"), the N/M denominator logic and zero-match interpretation table in step 4, and the TD-CHANGE-IMPACT gate with APPROVE/CONCERNS/REJECT/NOT ASSESSED verdicts and recovery loops ("re-analyze the impact before continuing", "supply it and re-run the gate").	5 / 5
Progressive Disclosure	Policy detail (automation modes, workflow tiers, director gates) is correctly delegated to clearly signaled one-level-deep project docs (".claude/docs/automation-modes.md", ".claude/docs/workflow-modes.md", ".claude/docs/director-gates/td-change-impact.md") rather than inlined. Not 5: the ~330-line body carries substantial inline material (report templates, dialog scripts, interpretation tables) that a bundle reference file could offload; not 3 since the inline/external split is mostly well-judged and navigation is easy.	4 / 5
	Total	18 / 20 Passed

Description

78%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.

A strong, terse description that names concrete actions, an explicit trigger, and a distinct niche. Its main weakness is compression: the trigger clause and keyword set could be expanded with natural synonyms ("design change", "supersede") and a fuller "Use when..." phrasing.

Suggestions

Add natural trigger synonyms and variations (e.g., "design doc change", "stale/superseded ADRs") so users phrasing the need differently still match.

Make the "when" clause more explicit, e.g., "Use when a GDD is edited and the ADRs written against it may be stale" instead of the compressed leading clause "A GDD changed".

Mention the classification outcome (Still Valid / Needs Review / Superseded) to round out the action coverage.

Dimension	Reasoning	Score
Specificity	Names several concrete actions — "scan ADRs and the traceability index for now-stale architectural decisions", "Impact report, guides resolution" — covering the core workflow. Not 5: it omits key actions the skill performs (classifying ADRs as valid/superseded, updating the traceability index).	4 / 5
Completeness	Both "what" (scan for stale decisions, produce impact report, guide resolution) and "when" are present — "A GDD changed" is an explicit trigger clause. Not 5: the "when" is a terse leading clause rather than a fuller explicit trigger phrase (e.g., "Use when a GDD changes and its ADRs may be stale").	4 / 5
Trigger Term Quality	Good domain keyword coverage ("GDD", "ADRs", "traceability index", "architectural decisions", "impact report") that users of this workflow would naturally say. Not 5: missing common variations like "design change", "stale/superseded decisions", or "update ADRs"; not 3 since coverage is genuinely good, not partial.	4 / 5
Distinctiveness Conflict Risk	Clear niche — propagating GDD changes to ADR staleness via a traceability index — with distinct domain triggers (GDD/ADR) unlikely to fire for any other skill.	5 / 5
	Total	17 / 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/propagate-design-change/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.