grc-planner

Pass

Audited by Gen Agent Trust Hub on Apr 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script query_grc_blocks.py with arguments to query block definitions. This is a functional requirement for the skill's purpose.
  • [PROMPT_INJECTION]: The skill processes external .grc files, creating an indirect prompt injection surface. Evidence: 1. Ingestion points: .grc file artifacts (Step 1). 2. Boundary markers: Interactive readiness check and structured markdown template. 3. Capability inventory: File creation and local shell script execution. 4. Sanitization: No explicit content sanitization for input artifacts is defined.
  • [SAFE]: The identified command execution and file access patterns (specifically reading from ~/Documents/grc-block-query/db) are localized and necessary for the primary function of GNU Radio flowgraph engineering.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 27, 2026, 03:44 PM