parallel-task

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection vulnerability surface because it processes external markdown plan files and interpolates user-controlled task descriptions directly into subagent prompt templates.
  • Ingestion points: Processes workspace plan files (e.g., plan.md, auth-plan.md) to extract task details in SKILL.md.
  • Boundary markers: Absent; there are no specific encapsulation tags or instructions to prevent text inside the task content from influencing subagent behavior.
  • Capability inventory: Spawns parallel subagents, executes test commands/validation steps, and performs file modifications and local git commits as described in SKILL.md.
  • Sanitization: Absent; no escaping, filtering, or validation is performed on the extracted task fields before embedding them into the subagent prompt template.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:00 AM
Security Audit — agent-trust-hub — parallel-task