Conciseness

The body carries no padding and no explanations of concepts Claude already knows — every sentence is a project-specific constraint — but it is ~300 lines with run-on paragraph walls (e.g., the 38-line paragraph in "What Gets Resolves") and repeated clauses ("resolve each value once" / "resolve once" / "resolve exactly once" reappear across the cookie, token-refresh, documents, and OAuth sections) that could be tightened. Matches anchor 3; not 2 because nothing is filler, not 4 because the prose needs restructuring rather than minor trims.

Actionability

Concrete executable elements abound: caddyfile placeholder examples, the exact lookup syntax "secrets:<secret_id>:<key>", named fixture files ("<prefix>.env", "<prefix>_resolved.json"), named tests, and a specific pipeline ("Raw encoded directive fields usually need `cfgutil.DecodeArgs`, replacement of each argument, lossless encoding, and validation"). Not 5 because there is no worked end-to-end fixture example showing the transform from .Caddyfile/.env to resolved JSON.

Workflow Clarity

"When adding a new placeholder-bearing field, check the authcrunch struct and validation path first" establishes a clear sequence (check struct → decode → replace → re-encode → validate → extend test coverage), and validation checkpoints with fail-closed semantics and named tests ("the old deployment still authorizes requests") are emphasized throughout. Not 5 because the full workflow is never assembled into one coherent ordered flow — the reader must piece it together from scattered sections.

Progressive Disclosure

Section headers are clear and the cross-references to sibling skills are well signaled and one level deep (e.g., "See [persistent runtime state](../configuration-state/SKILL.md)"), but the skill has no bundle files of its own and ~200 lines of per-field resolution catalog are inlined in "What Gets Resolved" — exactly the material that belongs in a separate reference file. Matches anchor 3 (content that should be separate is inline); not 4 because the core detail is not split out of SKILL.md.

Specificity

"Configure and validate Caddy runtime placeholders, secret lookups, encoded instructions, and resolved fixtures" plus "determine which fields resolve and preserve exact values" lists several concrete capabilities. Not 5 because the items are object nouns more than distinct actions and "resolved fixtures" is vague; clearly above 3 because coverage goes beyond 1-2 actions.

Completeness

The "what" is concretely stated, but there is no "when" clause: "Use to determine which fields resolve and preserve exact values" describes purpose, not a trigger situation (e.g., "Use when generating Caddyfiles with runtime placeholders"). Per the judging guideline, a missing explicit trigger caps completeness at 3; not 2 because the "what" is specific rather than vague.

Trigger Term Quality

Natural terms like "Caddy", "placeholders", and "secret lookups" are present, but common user variations ("{env.*}", "environment variables", "Caddyfile") are missing, and phrases like "resolved fixtures" and "manager blocks belong to secrets" are internal jargon rather than what a user would say. Matches anchor 3 (some relevant keywords, missing variations), not 4 because keyword coverage leans on project vocabulary rather than natural phrasing.

Distinctiveness Conflict Risk

The niche (Caddy runtime placeholder/secret resolution) has mostly distinct triggers, though vocabulary like "placeholders" and "secret lookups" overlaps with closely related sibling skills (configuration-secrets, configuration-state, cookies). Minor overlap risk with closely related skills matches anchor 4; not 5 because the description does not sharply separate it from that skill family.