Content
75%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A well-structured, actionable skill body with dual-language verify code, a clear thin-ping model, and dense gotchas, supported by real one-level reference files. Its main gaps are a payloads-fetch loop left as prose and example/ links that do not resolve in the bundle.
Suggestions
Add an executable payloads-fetch loop snippet (cursor pagination while mightHaveMore) alongside the verify functions so the common handler path is fully copy-paste ready.
Either include the examples/express, examples/nextjs, and examples/fastapi directories in the bundle or remove/qualify those inline links so navigation resolves.
Trim the Attribution block and the Related Skills / Recommended-skill link lists (or move them to a single references file) to recover tokens without losing the core guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence (no padding about what HMAC or webhooks are), but the Attribution block, the 'Recommended: webhook-handler-patterns' section with four external links, and the six-entry Related Skills list add tokens beyond the core task; not a 5 because those navigation/link sections could be trimmed, and not a 3 because the technical content itself is efficient. | 4 / 5 |
Actionability | Provides complete copy-paste verify() functions in both Node and Python, concrete env vars, a dev command, and the exact payloads endpoint with cursor/limit semantics; not a 5 because the payloads fetch loop is described in prose ('loop while true; max limit is 50') rather than as executable code, leaving a minor gap in the most common handler path. | 4 / 5 |
Workflow Clarity | The two-step thin-ping model is clearly sequenced (respond 200/204 within 25s, then fetch payloads by cursor) and signature verification acts as the validation checkpoint, with retry/rate-limit recovery in Gotchas; not a 5 because the full receive→verify→fetch→handle flow is distributed across sections rather than presented as one numbered workflow with explicit inline checkpoints. | 4 / 5 |
Progressive Disclosure | SKILL.md stays an overview and pushes detail to well-signaled one-level-deep references (references/overview.md, setup.md, verification.md — all present in the bundle); not a 5 because the body also references examples/express/, examples/nextjs/, and examples/fastapi/ which do not exist in the bundle, creating broken navigation paths. | 4 / 5 |
Total | 16 / 20 Passed |