Content
96%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.
An exceptionally strong, dense playbook: executable detection and exploitation commands, a technique matrix, explicit validation checkpoints, and an OPSEC/decision gate that gates live exploitation on defender notification. The only structural flaw is the dangling 'finding-protocol' reference, which points outside the skill to an undefined target.
Suggestions
Resolve the 'finding-protocol' reference — either point to a real file in the bundle (e.g., references/finding-protocol.md) or inline the escalation/downgrade instructions so the final workflow step is self-contained.
Consider adding a fat-GET and X-Original-URL routing-override PoC alongside the three existing ones, since both are called out as significant techniques in the matrix but lack executable examples.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean playbook with zero padding: the only background material (the unkeyed-input definition and the cache-poisoning vs cache-deception disambiguation) is domain-specific and load-bearing, not concepts Claude already knows. Every section (detection, matrix, PoCs, chains, tools, OPSEC, gate) earns its tokens. | 5 / 5 |
Actionability | Fully executable bash: cache-status fingerprinting, a header-permutation loop with reflection greps, a three-step poisoning-confirmation procedure, and copy-paste PoCs for stored-XSS, forced redirect, and param cloaking. The '<TARGET>' placeholder is the standard template pattern for an authorized-target security skill, and the examples cover the common cases. | 5 / 5 |
Workflow Clarity | Clear numbered sequence (detect → confirm → exploit → chains → tools → OPSEC) with explicit validation at each stage: the second-request-without-header confirmation, cache-HIT verification in each PoC, and a Decision Gate checklist requiring defender notification before live-cache exploitation. The destructive/batch cap does not apply because validation checkpoints are present. | 5 / 5 |
Progressive Disclosure | Well-organized sections with content appropriately sized for a single-file skill (no bundle files exist, none needed). However, the closing gate says 'escalate per finding-protocol', a reference that does not exist in this bundle — no references/, scripts/, or sibling file resolves it — leaving navigation dangling at the final step. | 4 / 5 |
Total | 19 / 20 Passed |