Content
82%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 strong reference-style skill body: nearly everything shown is executable and Deck-specific, verification includes genuine fail-closed error handling, and progressive disclosure mostly offloads bulk detail to three well-signaled reference files. The main costs are a padded tail of cross-promotional link sections, an implicit rather than explicit end-to-end workflow, and references to examples/ directories that are absent from the bundle.
Suggestions
Trim the 'Related Skills' list (12 external links), the 'Recommended' section, and the 'Attribution' section to a few lines or move them to a reference file — roughly 20 lines of the body are cross-promotion rather than skill content.
Add a short numbered setup workflow (generate secret → create destination → return 2xx to activate → verify deliveries) with the pending_verification checkpoint inline, instead of leaving the sequence implicit across topic sections and in references/setup.md.
Either include the examples/express, examples/nextjs, and examples/fastapi handler directories in the bundle or remove/replace those links with pointers to content that exists (e.g., a complete handler in references/verification.md).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The instructional core is lean and assumes competence — no "what is a webhook" padding, and sections like Verification and Delivery Semantics carry only Deck-specific facts ("5-minute timestamp tolerance", "previous one valid 24h after rotation", "base 2, starting at 30s"). However, the trailing "Related Skills" section (12 external skill links), the "Recommended: webhook-handler-patterns" section with 4 more links, and the "Attribution" section are tokens that don't serve the skill's task. Not 3: this is link-list padding rather than unnecessary explanation, and the actual guidance is trim; not 5: those trailing sections keep it from "every token earns its place". | 4 / 5 |
Actionability | Fully executable, copy-paste-ready guidance covering the common cases: a node one-liner to generate the secret, a complete curl call to create the destination, a working fail-closed JavaScript verifier with dual-secret rotation handling and explicit 400/500 outcomes, the Python equivalent, the interaction-answer endpoint ("POST /v2/task-runs/{data.id}/interaction"), the failure-inspection endpoint, and a hookdeck CLI command. Matches the score-5 anchor; a 4 would require minor gaps, and none of the shown code is pseudocode or incomplete. | 5 / 5 |
Workflow Clarity | The setup-to-handling flow is clear in section order (generate secret → create destination → 2xx activates it → verify each delivery → handle idempotently), and verification has real checkpoints ("fail closed → 500", "→ 400 Invalid signature", try both secrets then rethrow the last error). Not 5: the body never presents this as an explicit sequenced workflow with validation feedback loops — the numbered steps live in references/setup.md, and the reader must assemble the sequence across topic sections. Not 3: the checkpoints that are present (activation via 2xx, fail-closed behavior, failure inspection via GET /event-deliveries?status=failure, dedupe guidance) are explicit, not implicit. | 4 / 5 |
Progressive Disclosure | Good structure against the actual bundle: the three real reference files (references/overview.md, setup.md, verification.md) are each linked inline where relevant and listed with one-line descriptions in a Reference Materials section, one level deep, and bulk detail (full event catalog, manual verification) is correctly kept out of SKILL.md. Not 5: the body points to "examples/express/", "examples/nextjs/", and "examples/fastapi/" for complete handlers, but no examples/ directory exists in the bundle — dangling paths that break easy navigation. Not 3: references are clearly signaled and content is appropriately split, well above the 'buried references' anchor. | 4 / 5 |
Total | 17 / 20 Passed |