handwerksrolle-hwo
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data—specifically employment dates and activity descriptions—which are then used to populate arguments for shell commands. If an attacker provides strings containing command injection payloads instead of dates, the calculation script's invocation could be compromised.
- Ingestion points: User-provided activity descriptions, qualifications, and employment history dates defined in the 'Eingaben' section of SKILL.md.
- Boundary markers: Absent. There are no instructions for the agent to wrap user data in delimiters or to ignore embedded instructions when formatting the calculation command.
- Capability inventory: The agent is instructed to execute local shell commands using the
python -m scripts.legal_calc.climodule. - Sanitization: No mention of validating date formats or escaping user-provided content before passing it to the CLI tool.
- [COMMAND_EXECUTION]: The skill explicitly directs the agent to interact with the underlying system via a project-local script. While the script itself is local, the execution pattern is a bridge between untrusted user input and shell-level operations.
Audit Metadata