mission-long-task
Pass
Audited by Gen Agent Trust Hub on May 27, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection (Category 8).
- Ingestion points: The skill ingests complex task descriptions from users during Phase 1 (SKILL.md).
- Boundary markers: Absent; there are no instructions to use delimiters or ignore instructions within the user-provided data.
- Capability inventory: The skill has the capability to write to the filesystem (creating .mission/ directories and CSV files) and hand off tasks to the mission-csv-execute tool (SKILL.md).
- Sanitization: Absent; the instructions do not include steps to sanitize or validate user input before it is used to generate the mission CSV.
- [COMMAND_EXECUTION]: The skill utilizes dynamic execution (Category 10) by generating a mission-specific CSV file at runtime that dictates the agent's future actions via mission-csv-execute. While this is the intended functionality, the lack of input validation makes this execution path dependent on potentially untrusted input.
Audit Metadata