Content
76%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 lean, concrete attack catalog with mostly copy-paste-ready payloads and good section organization. Its weaknesses are the absent validation/decision checkpoints (how to confirm a WSDL parsed, how to read SOAP fault responses, which attack to try first) and a few incomplete examples (the @ws-attack.xml payload file, the WS-Addressing fragment).
Suggestions
Add an explicit attack sequence with checkpoints, e.g. '1. Fetch and parse the WSDL, 2. List operations/parameters from the XSD types, 3. Pick an attack from the catalog, 4. Inspect the SOAP fault code/body to confirm or rule out each vulnerability'.
Provide the complete @ws-attack.xml UsernameToken payload inline (or as a bundle file) and a full WS-Addressing replay request instead of a one-line fragment.
Add brief response-interpretation guidance for each attack (what a successful XXE, UsernameToken hit, or schema-bypass response looks like versus a generic SOAP fault).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes competence: a one-line context sentence, then immediately concrete commands and payloads with only functional comments (e.g. "# WSDL endpoints — try ?wsdl on any .asmx (.NET) / .svc (WCF) / Spring-WS endpoint"). No concepts are re-explained and every section earns its tokens. | 5 / 5 |
Actionability | Mostly executable: the XXE curl with full SOAP envelope, Billion-Laughs payload, SOAPAction routing curl, and discovery commands are copy-paste ready. Minor gaps remain — "curl -sk -X POST -d @ws-attack.xml" references a payload file that is never provided, the WS-Addressing replay shows only a fragment instead of a full request, and the UsernameToken brute is a comment ("Brute: cycle a username dict, observe response time/error class") rather than a runnable command. | 4 / 5 |
Workflow Clarity | The sequence is implicit (Discovery section first, then the attack catalog, then tooling/OPSEC) but there is no explicit ordering guidance, no validation checkpoints (e.g. confirm the WSDL parsed and identify the real operation names before attacking), and no feedback loop for interpreting SOAP faults versus successful injection. This matches 'sequence present but checkpoints missing or implicit'. | 3 / 5 |
Progressive Disclosure | Well-organized single-file structure with clear sections (Discovery → attack catalog → Tooling → OPSEC → References) and a clearly signaled one-level-deep deferral to the SAML skill for XSW patterns. At ~110 inline lines it exceeds the 'under 50 lines' exception, and the tooling/catalog detail could plausibly live in a reference file, so it falls just short of the top anchor. | 4 / 5 |
Total | 16 / 20 Passed |