Content
77%Reviews 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 hunting playbook with strong validation discipline. Its main weaknesses are duplicated content across sections and a monolithic structure that keeps reference material inline rather than in separate files.
Suggestions
De-duplicate: keep the OOB/blind confirmation logic and severity rubric in one place ('Validation Discipline') and have the earlier 'OOB Confirmation Gate' section cross-reference it instead of restating it.
Split the Bypass Table, Named CVEs, and Sensitive Files lists into references/ files (e.g. bypass-table.md, cves.md, sensitive-files.md) and link to them one level deep from SKILL.md to improve progressive disclosure.
Trim the conceptual framing in 'OOB / Blind-LFI Confirmation Gate' (what 'blind' means) — a skilled model knows this; keep only the tactical confirmation criteria.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly lean command/table reference, but material is duplicated across sections (OOB confirmation appears in both 'OOB Confirmation Gate' and 'Validation Discipline'; severity in both 'Crown Jewels' and 'Validation Discipline'), so not every token earns its place. | 2 / 3 |
Actionability | Fully executable: concrete ffuf/wfuzz/dotdotpwn commands, exact php://filter payloads, and copy-paste-ready bash blocks for every phase. | 3 / 3 |
Workflow Clarity | A clear 9-phase sequence with explicit validation gates and retract rules ('Claim LFI only after...'; triple-confirm timing deltas), giving feedback loops for the batch/automation steps. | 3 / 3 |
Progressive Disclosure | Well-organized with headers and tables, but it is a single 262-line monolith; the bypass table, CVE list, and sensitive-file catalog are inline reference content that would score higher split into one-level-deep reference files. | 2 / 3 |
Total | 10 / 12 Passed |