skills/seanmars/mreq/mreq-grill/Gen Agent Trust Hub

mreq-grill

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from task documents and source code to identify planning assumptions. 1. Ingestion points: Reads .mreq/ documents and cited code in Step 2. 2. Boundary markers: No explicit markers are used to separate document content from agent instructions. 3. Capability inventory: Execution of local Python scripts, filesystem edits, and skill invocation (mreq-apply). 4. Sanitization: Document content is analyzed directly without validation.
  • [COMMAND_EXECUTION]: The skill executes a local utility script at ../mreq-shared/scripts/mreq.py via shell commands (uv run, python, or python3) to list and update document metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 04:17 PM