Content
92%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.
The body is a well-structured, highly actionable multi-phase workflow with strong validation and feedback loops, and it cleanly separates overview from one-level-deep references. The only weakness is mild redundancy that could be tightened to improve token efficiency.
Suggestions
Remove the Invocation trigger-phrase list from the body since it duplicates the frontmatter description verbatim; reference 'the triggers in the description' instead.
Consolidate the Agent Integrity Rules block with the per-phase rate-limit/retry notes to avoid restating the same discipline twice.
Trim the Anti-Patterns To Reject list to the few items not already enforced inline (e.g., dedup and empty-file rules are covered in the Error Handling table).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Largely efficient and assumes Claude's competence (no generic explanations of what Reddit/HN are), but there is redundancy — the Invocation section re-lists trigger phrases already in the frontmatter, Agent Integrity Rules are restated at phase level, and the 10-item Anti-Patterns list could be trimmed — keeping it just below the lean/5 anchor. | 4 / 5 |
Actionability | Fully executable guidance throughout: concrete endpoints ('reddit.com/search.json', 'hn.algolia.com/api/v1/'), exact query parameters, citation_tracker.py invocations with flags, and a copy-paste bash block for the X export importer. Not below 5: minor placeholders like <topic> are appropriate parameterization, not gaps. | 5 / 5 |
Workflow Clarity | A clear dependency-ordered sequence (Phase 0 intake → Pre-flight → Phases 1–4 in parallel → Synthesis → Output) with explicit validation checkpoints (three-count audit log, citation discipline, retry policy) and feedback loops (failure → 3s wait → retry once → fallback → 3-consecutive-failures stop), satisfying the batch-operation validation requirement. | 5 / 5 |
Progressive Disclosure | SKILL.md is an overview with well-signaled, one-level-deep references to real files (references/research_pack_conventions.md, parallel_execution_discipline.md, cross_platform_synthesis.md) and a clear scripts/ table; the inline Agent Integrity Rules summary correctly defers the canon to a reference rather than nesting. | 5 / 5 |
Total | 19 / 20 Passed |