api-exploit-prover
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [NO_CODE]: The skill is composed entirely of markdown instructions defining a methodology for API vulnerability confirmation. It does not include any scripts, binaries, or executable configuration files.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing untrusted data to perform potentially sensitive security testing actions, creating a surface for indirect prompt injection.
- Ingestion points: The skill takes external inputs such as
candidate_findingsandtest_data_or_seed_objectsinSKILL.md. - Boundary markers: There are no specific delimiters or instructions to ignore potential commands embedded within these untrusted inputs defined in the provided methodology.
- Capability inventory: The execution workflow (Phases 1-3) directs the agent to perform request replays, state changes, and unauthorized data access tests based on the provided leads.
- Sanitization: No specific sanitization, escaping, or validation steps for external content are specified in the instructions to prevent malicious payloads in findings from influencing agent behavior.
Audit Metadata