repo-intake-and-plan

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted data from external repositories to inform the agent's planning phase, creating a surface for indirect prompt injection.
  • Ingestion points: The scripts scripts/extract_commands.py and scripts/scan_repo.py read content directly from the target repository's README files and file system structure.
  • Boundary markers: There are no specific boundary markers or instructions to the agent to disregard potential instructions embedded within the untrusted text being processed.
  • Capability inventory: While the SKILL.md explicitly states the skill does not execute commands, its output (a reproduction plan) is intended to guide a main skill that likely has environment setup and command execution capabilities.
  • Sanitization: The extraction logic classifies commands based on keywords but does not perform security validation or sanitization on the extracted command strings to prevent shell-based attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 02:58 PM