Content
63%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 dense, well-sectioned body that is strong on executable tool-call examples and includes a rigorous signed-release verification procedure, but it is held back by duplicated 'Common Mistakes' entries, marketing claims, an oversized inlined bash script, and dangling INSTALL.md/docs/ references that don't exist in the bundle. Trimming the duplication and moving the verification script to a bundled script or reference file would improve both conciseness and progressive disclosure.
Suggestions
Merge the near-duplicate 'Ignoring exploitability context' and 'Skipping critical severity' Common Mistakes entries into one, and consider dropping the Quick Reference rows that merely repeat the preceding tool sections.
Move the ~70-line release verification script into a bundled scripts/ or references/ file and reference it with a one-line pointer, keeping SKILL.md as the overview.
Ensure ./INSTALL.md and ./docs/ actually ship in the bundle (they are absent here), or remove the links so navigation doesn't dead-end.
Add a short remediation step to the audit workflows (what to do after 'Urgent advisories found!') to close the workflow feedback loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly tight, executable code, but 'Common Mistakes' #2 ('Ignoring exploitability context') and #3 ('Skipping critical severity') are near-duplicates with identical DO-blocks, the Quick Reference table repeats the preceding tool sections, and claims like 'Zero false positives (curated feed only)' add marketing padding — matching 'mostly efficient but includes some unnecessary explanation or could be tightened'. It is above anchor 2 because there is no explanation of concepts Claude already knows; the hardcoded VERSION="0.0.10" is time-sensitive and not in a deprecated/old-patterns section, which also costs it anchor 4. | 3 / 5 |
Actionability | Concrete TypeScript tool calls with real parameter names (skillName, installRoot, severity, exploitabilityScore), a Quick Reference parameter table, and a fully copy-paste-ready bash verification script with checksums and signature checks satisfy 'mostly executable guidance; concrete code or commands with minor gaps'. It falls short of anchor 5 because the pattern examples depend on undefined helper functions (installSkill, showSecurityWarning, getUserConfirmation, notifyNow) that don't exist anywhere in the bundle, so those patterns cannot run as written. | 4 / 5 |
Workflow Clarity | Pattern 1 gives an explicit check-then-act sequence with a user-confirmation branch, and the release-artifact verification section has explicit validation checkpoints with error exits ('if [ "$ACTUAL_PUBKEY_SHA256" != ... ] exit 1') and the closing rule 'Only install or extract the archive after this verification succeeds' — clear sequence with most checkpoints present. It is not anchor 5 because the audit workflows lack remediation loops: what to do after 'Urgent advisories found!' or a critical match is never sequenced beyond alerting. | 4 / 5 |
Progressive Disclosure | Section structure is good, but scored against the actual bundle (no references/, scripts/, or assets/ directories exist): the ~70-line release-verification bash script is inlined content that clearly belongs in a separate script or reference file, and the only external pointers — [INSTALL.md](./INSTALL.md) and [docs/](./docs/) — reference files not present in the bundle. This matches 'some structure but could be better organized; content that should be separate is inline' with references that are not usable as-is. | 3 / 5 |
Total | 14 / 20 Passed |