t-plan
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection attacks where malicious instructions embedded in the codebase could influence the behavior of subagents.\n
- Ingestion points: Subagents are directed to read files such as
intent.md,explore.md, and other project files during theEXPLOREandVALIDATEphases.\n - Boundary markers: The instructions provide context labels (e.g.,
**Context**: Read ...) but lack explicit delimiters or instructions for subagents to ignore potentially malicious commands within the files being analyzed.\n - Capability inventory: The system allows the orchestrator to execute shell commands (
bash) and spawn further agents (Agent), providing a path for injected instructions to escalate actions.\n - Sanitization: There is no evidence of filtering, escaping, or validation of the content read from external files before it is processed by the subagents.
Audit Metadata