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, security-conscious skill body: phases are clearly gated by authorization, most of the core workflow is command-driven, and boundaries are explicit. The main gaps are a handful of high-level sections without concrete commands (service enumeration, fingerprinting, code-repo searching) and the absence of validation/feedback checkpoints after active steps.
Suggestions
Add concrete commands or named tools to the vague sections — service enumeration (e.g., an nmap NSE or version-probe example), technology fingerprinting (e.g., a Wappalyzer/whatweb invocation), and GitHub/GitLab dorking (e.g., specific search query patterns) — to lift actionability from mostly-executable to fully executable.
Insert validation checkpoints in the workflow: re-confirm scope immediately before Phase 2, and a step to verify/handle failed or filtered scan results before enumerating services, which would also strengthen the feedback-loop quality of workflow_clarity.
Tighten small filler phrases (e.g., "Gather information without touching the target directly", "Correlate all findings.") that add tokens without adding instruction, nudging conciseness toward the lean anchor.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient — commands are given directly (dig, whois, nmap, crt.sh pipeline) with no tutorials on what these tools are and no concept explanations Claude already knows. Minor instances could be trimmed (e.g., "Gather information without touching the target directly", "Correlate all findings"), placing it just below the lean-and-efficient anchor 5 and well above anchor 3's unnecessary-explanation threshold. | 4 / 5 |
Actionability | Most guidance is executable: `dig any $ARGUMENTS`, `dig axfr @ns-server $ARGUMENTS`, the crt.sh curl|jq pipeline, `whois $ARGUMENTS`, `nmap -sC -sV -oN scan-results.txt $ARGUMENTS`, named tools (gobuster/feroxbuster/dirsearch, testssl.sh/sslyze), and concrete API paths (/api/, /v1/, /graphql, /swagger.json). Gaps remain in service enumeration ("probe for version info and default configurations"), technology fingerprinting, and GitHub/GitLab searching, which stay high-level — mostly executable with minor gaps rather than the incomplete-guidance anchor 3. | 4 / 5 |
Workflow Clarity | A clear three-phase sequence (Passive → Active → Analysis) with an explicit authorization gate before any commands ("Before running any commands, confirm: 1... 2..."), an explicit-authorization marker gating Phase 2, prioritization criteria, and a structured output template. It falls short of anchor 5 because there are no post-scan validation/feedback loops (e.g., re-verify scope before active steps, handle failed or filtered scans), but checkpoints are explicit rather than implicit, ruling out anchor 3. | 4 / 5 |
Progressive Disclosure | No bundle files exist and the ~107-line body is well organized into clearly headed sections, with sibling-skill cross-references clearly signaled and a References section pointing to external standards (PTES, OWASP, jhaddix/tbhm) rather than nested files. Good structure with minor organization gaps — the methodology detail could arguably live in a reference file — so anchor 4 rather than the simple-skill anchor 5. | 4 / 5 |
Total | 16 / 20 Passed |