agent-task-splitter
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting user-supplied goals and success criteria for decomposition into task packets.
- Ingestion points: The 'Inputs' section in SKILL.md (Goal, authorized scope, success criteria).
- Boundary markers: Absent; the skill does not explicitly instruct the agent to use delimiters or warnings when interpolating user goals into the generated .ai/task_*.md files.
- Capability inventory: The skill performs file writing operations to create .coord/plan.yml and .ai/task_*.md files (SKILL.md).
- Sanitization: Absent; no instructions for escaping or filtering user-provided goal text before writing to task files.
- [COMMAND_EXECUTION]: The workflow involves generating instructions for a host to execute the agent-collab CLI tool to perform policy evaluations (SKILL.md).
Audit Metadata