uipath-planner
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill represents an orchestration layer that processes untrusted process-knowledge sources to generate technical designs and downstream agent instructions.
- Ingestion points: Untrusted data enters the agent context through the Entry Guard in "SKILL.md" and the ingestion procedures for PDFs, Word documents, and transcripts described in "references/pdd-analysis-guide.md".
- Boundary markers: The skill's instructions do not specify the use of delimiters or "ignore embedded instructions" warnings when processing process descriptions provided by users or external documents.
- Capability inventory: The skill possesses powerful capabilities, including creating and updating tasks for other specialist agents via "TaskCreate" and "TaskUpdate", and executing local shell commands via the Bash tool (e.g., using pandoc via scripts in the "scripts/" directory).
- Sanitization: There is no evidence of sanitization, escaping, or filtering of the content extracted from external process documents before it is interpolated into Solution Design Documents or Skill prompts for implementation tasks.
Audit Metadata