task-breakdown

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project documentation to generate task boards, creating a surface where malicious instructions in those documents could influence the agent.
  • Ingestion points: docs/business/user-story.md, docs/business/acceptance-criteria.md, and docs/business/acceptance-criteria-breakdown/*.md (referenced in SKILL.md).
  • Boundary markers: Absent; the skill does not use delimiters to wrap the ingested content.
  • Capability inventory: File-write access to docs/task-breakdown/ and shell command execution via local scripts (documented in SKILL.md).
  • Sanitization: Absent; documentation strings are interpolated into the output files.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to run local validation and formatting scripts.
  • Evidence: SKILL.md contains instructions to execute python3 <this skill's dir>/scripts/recompute_summary.py and python3 <this skill's dir>/scripts/check_ac_refs.py to maintain task board integrity.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 04:46 AM