cowork-sessions

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill architecture is susceptible to indirect prompt injection through its session orchestration workflow.
  • Ingestion points: The skill processes 'task descriptions' during the INITIALIZING phase to determine subtasks and execution order (SKILL.md).
  • Boundary markers: There are no specified delimiters or instructions defined in the documentation to isolate or ignore potentially malicious instructions embedded within these external task descriptions.
  • Capability inventory: The skill is configured with 'Bash' and 'Task' tools, the latter of which is used to spawn sub-agents for executing subtasks derived from the potentially untrusted input.
  • Sanitization: The documentation does not specify any validation, filtering, or sanitization steps for the 'task description' input before it influences the planning and agent dispatching stages.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 11:55 PM