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 highly actionable, well-sequenced offensive-testing workflow with strong executable commands, explicit validation, and a useful scenario/pitfalls section. Its weaknesses are padded concept/tool explanations Claude already knows and a monolithic structure that ignores the skill's own bundle files.
Suggestions
Remove or drastically trim the 'Key Concepts' and 'Tools & Systems' sections — they re-explain concepts (HSTS, ARP poisoning, what Wireshark is) that Claude already knows.
Link the existing bundle files from the body (e.g., 'Full command reference: see references/api-reference.md; automation script: scripts/agent.py') and move per-tool detail there so the SKILL.md is a lean overview.
Convert comment-only validation steps in Steps 5 and 6 (Cisco DAI log check, target-side ARP cache refresh) into executable commands or a short verification checklist with a retry/recovery step.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The bulk of the body is lean, executable command blocks, but the 'Key Concepts' table (MITM, SSL Stripping, HSTS, ARP Cache Poisoning) and 'Tools & Systems' section explain concepts and tools Claude already knows, which is unnecessary padding. | 3 / 5 |
Actionability | Mostly copy-paste-ready guidance — complete bettercap/ettercap/mitmproxy command sequences, a fully written Python mitmproxy script, and concrete validation commands — with minor gaps where several Step 5/6 checks are comment-only placeholders (e.g., '# On Cisco switch: show ip arp inspection log'). | 4 / 5 |
Workflow Clarity | Six clearly sequenced steps with an explicit validation step (Step 5) and a cleanup/restore step (Step 6), so the destructive-operation cap does not apply; however, there is no explicit error-recovery feedback loop (e.g., what to do if the target loses connectivity mid-test) and some checkpoints are comments rather than commands. | 4 / 5 |
Progressive Disclosure | Section headers are clear, but the body never references the existing bundle files (references/api-reference.md, scripts/agent.py), leaving them orphaned, and reference-worthy content (Key Concepts, Tools & Systems, per-tool detail) is inlined in the SKILL.md rather than split out. | 3 / 5 |
Total | 14 / 20 Passed |