Conciseness

The body is code-dominant with terse rules ("Prefer short, modern, copyable patterns. Do not write tutorial prose unless the user explicitly asks for explanation") and explains no concepts Claude already knows. It falls short of lean because core rules repeat across three sections (e.g., node-before-execute appears in "Use These Defaults", "Critical Rules", and "Do Not Generate") and the 11-diagram Flow Gallery plus dense runtime-hook resolution rules carry more detail than codegen strictly needs — anchor 4, not 3, since there is no padding or tutorial prose.

Actionability

Nearly every section is complete, copy-paste TypeScript: the Canonical Pattern runs end-to-end ("const result = await wf.forward(llm, { userInput: 'Hello world' })"), each feature (branch, while, feedback, parallel, derive, mermaid) has a full executable example, and common errors map to concrete fixes ("\"Node 'x' not found\" -- define .node() before .execute()"). This matches the fully-executable, common-cases-covered anchor.

Workflow Clarity

The codegen workflow is unambiguous: defaults, critical rules, a canonical pattern, per-feature patterns, an error-to-fix table, and a "Do Not Generate" checklist serve as explicit checkpoints, and error-recovery guidance is concrete. As a single-purpose skill whose single action is unambiguous — with explicit error-feedback mapping — it meets the anchor-5 bar; there are no destructive or batch operations that would require additional validation.

Progressive Disclosure

Structure is good: clear section headers, logical grouping, related domains deferred rather than inlined ("full modifier table in the ax-signature skill", "Use `ax-mcp` for MCP client construction"), and full examples linked externally. It is not 5 because no bundle files exist at all — the ~130-line Flow Gallery, the mermaid dialect spec, runtime hooks, and event-triggered flows are advanced material inlined in an already ~750-line body that would fit dedicated reference files; it is not 3 because organization and signaling are strong throughout.

Specificity

"This skill helps an LLM generate correct AxFlow workflow code using @ax-llm/ax" names the domain and one concrete action, but never enumerates the skill's other capabilities (conditional branching, while/feedback loops, mermaid-authored flows, multi-model routing) — anchor 3. Not 4 because only a single action is stated rather than several; not 2 because the stated action is concrete and library-specific, not generic. Voice is third person, so no penalty applies.

Completeness

It clearly answers both questions: "helps an LLM generate correct AxFlow workflow code using @ax-llm/ax" states what it does, and "Use when the user asks about flow(), AxFlow, workflow orchestration, …" explicitly states when with concrete trigger phrases — a direct match to the anchor-5 example pattern.

Trigger Term Quality

"flow(), AxFlow, workflow orchestration, parallel execution, DAG workflows, conditional routing, map/reduce patterns, or multi-node AI pipelines" gives good natural-keyword coverage a user of this library would plausibly say. Not 5 because terms for significant features the skill covers are missing — e.g., mermaid flow authoring, retry/feedback loops, and event-triggered flows.

Distinctiveness Conflict Risk

The AxFlow/@ax-llm/ax/flow() niche is clearly distinct, but several trigger phrases are generic workflow vocabulary — "DAG workflows", "map/reduce patterns", "workflow orchestration" — that could fire for orchestration questions unrelated to this library. Mostly distinct with minor overlap risk matches anchor 4; not 5 because of those generic terms, not 3 because the library-specific terms anchor it firmly.