Conciseness

The body re-explains fundamentals Claude already knows — "Resources are nouns (users, orders, products), not verbs", "GET: Retrieve resources (idempotent, safe)", "Queries for reading data / Mutations for modifying data", "Clients request exactly what they need" — and then inlines ~400 lines of pattern code (FastAPI pagination endpoint, ~90-line GraphQL schema, resolvers, DataLoaders) that largely duplicates references/rest-best-practices.md and references/graphql-schema-design.md. This is noticeably verbose with several padded sections, fitting the level-2 anchor rather than the minor-trimming of level 3.

Actionability

The patterns are conveyed through concrete, largely executable code: a complete FastAPI paginated list endpoint, ariadne resolver implementations with cursor pagination, and DataLoader classes with batch_load_fn logic. Minor gaps keep it below level 5: helpers like fetch_users, hash_password, decode_cursor and imports like Any and ValidationError are undefined, so examples are not fully copy-paste ready.

Workflow Clarity

There is no sequenced design or review process at all — the body is a catalog of concepts and patterns with no ordering (no 'design the schema, then the resources, then errors, then validate against the checklist' flow) and no validation checkpoints, even though the skill's stated use cases ('reviewing API specifications', 'establishing API design standards') are process-shaped and a checklist asset exists but is never wired into a workflow. This sits below level 3, which requires at least a present step sequence.

Progressive Disclosure

Section headers are clear and the Resources section signals bundle files with one-line descriptions, but content that should be separate is inlined: the REST/GraphQL pattern sections duplicate the 408-line rest-best-practices.md and 583-line graphql-schema-design.md. It cannot score 4 because navigation is partially broken — 3 of the 7 listed resources do not exist (references/api-versioning-strategies.md, assets/graphql-schema-template.graphql, scripts/openapi-generator.py, and there is no scripts/ directory at all).

Specificity

The description names the domain ("REST and GraphQL API design principles") and gestures at 1-2 actions ("build intuitive, scalable, and maintainable APIs"), but 'master principles' is abstract rather than a concrete capability, and 'delight developers' is buzzword fluff. It does not reach level 4 because it never lists several specific concrete actions, and it clears level 2 because the domain and a build action are explicitly stated.

Completeness

Both parts are present: the 'what' ("Master REST and GraphQL API design principles to build... APIs") and an explicit 'when' clause ("Use when designing new APIs, reviewing API specifications, or establishing API design standards"). It falls short of level 5 because the 'what' states outcomes rather than concrete capabilities, so the pairing is not as concretely matched as the anchor example.

Trigger Term Quality

Natural trigger phrases like "REST", "GraphQL", "API design", "designing new APIs", "reviewing API specifications", and "establishing API design standards" match what a user would actually say. Not level 5 because common variations such as 'endpoint design', 'API guidelines', 'versioning', or 'API documentation' are absent.

Distinctiveness Conflict Risk

The REST/GraphQL API-design framing carves a clear niche with distinct triggers that would not fire for unrelated skills. Not level 5 because the broad trigger 'designing new APIs' could overlap with general web-development or framework-specific skills.