skills/testland/qa/restler-fuzzing/Gen Agent Trust Hub

restler-fuzzing

Pass

Audited by Gen Agent Trust Hub on Oct 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to clone the RESTler source code from Microsoft's official GitHub repository (github.com/microsoft/restler-fuzzer) for installation and build purposes.
  • [COMMAND_EXECUTION]: The documentation includes standard shell, Python, and Docker commands required to compile the fuzzer grammar and execute the testing passes (compile, test, fuzz-lean, fuzz).
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data which could potentially contain malicious instructions.
    • Ingestion points: Reads external OpenAPI specifications (e.g., openapi.json) using the --api_spec flag in SKILL.md and the nightly CI workflow in references/triage-and-ci.md.
    • Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are provided for the input specification files.
    • Capability inventory: The skill utilizes Docker to run containers with network access for API fuzzing and file system access to write results to RestlerResults/.
    • Sanitization: Relies on the internal parsing and compilation logic of the Microsoft RESTler tool to handle the OpenAPI input.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 2, 2026, 10:10 AM