Content
88%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-built operational skill body: the phase workflow is precisely sequenced with disambiguation, handoff, and gap-labeling checkpoints, and the output format plus failure handling are copy-paste concrete. The minor costs are section-level redundancy (Tools Used, Related Skills, What this solves) and the absence of any bundled reference files to offload detail.
Suggestions
Trim or merge the 'Tools Used' section into the phase steps (it duplicates the frontmatter tools list) and collapse 'Related Skills and Operations' overlap with 'What this is not'.
Consider moving the seven-bucket classification definitions and the output-format template into a bundled reference file (e.g., references/lineage-buckets.md) so SKILL.md stays a lean overview, which would also make the external references verifiable within the bundle.
Trim the first paragraph of 'What this solves', which restates the frontmatter description almost verbatim.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The core sections (Phases, Output Format, Quality Rules, When it fails) are dense and every token earns its place, but 'What this solves' restates the description, 'Tools Used' duplicates the frontmatter tool list, and 'Related Skills' overlaps 'What this is not' — minor instances that could be trimmed. | 4 / 5 |
Actionability | As an instruction-only skill the guidance is fully concrete: exact tool per phase step (search, query, get_page, takes_search, find_contradictions), a copy-paste-ready output-format template, exact gap labels ('No clear evidence found', 'undated'), and specific error codes (page_not_found) with recovery actions. | 5 / 5 |
Workflow Clarity | Five clearly sequenced phases with explicit checkpoints: a disambiguation ask when multiple ideas share a phrase, a wrong-shape handoff to find_trajectory, gap labeling instead of fabricated narrative, and error-recovery feedback loops in 'When it fails'. The skill is read-only, so the destructive/batch validation cap does not apply. | 5 / 5 |
Progressive Disclosure | No bundle files exist (references/, scripts/, assets/ are absent), so all operational detail lives inline in a ~200-line body. The external references (../conventions/quality.md, ../conventions/agent-operator-protocol.md) are well-signaled and one level deep but point outside the skill directory and cannot be verified in this bundle — good structure with minor organization gaps. | 4 / 5 |
Total | 18 / 20 Passed |