design-workflow-graphs
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No prompt injection or behavior override patterns were found in the instructions.
- [SAFE]: No sensitive data exposure, hardcoded credentials, or network exfiltration patterns were identified.
- [SAFE]: The Python validation script (scripts/validate_graph_spec.py) uses only standard libraries and performs local structural and logical validation of JSON specifications without network access or unsafe file system operations.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied workflow requirements and existing graph specifications. The skill implements structured state definitions and a formal validation protocol to mitigate accidental misinterpretation. The provided tool performs local validation only.
- Ingestion points: Processes natural language requirements and graph specifications via user input and local files (SKILL.md, scripts/validate_graph_spec.py).
- Boundary markers: Encourages explicit state contracts, node responsibilities, and machine-readable JSON schemas (references/graph-model.md, assets/graph-spec.json).
- Capability inventory: Limited to local file reading and console output via the validation script (scripts/validate_graph_spec.py).
- Sanitization: The validation script enforces structural integrity, identifier constraints, and logical consistency of the graph topology (scripts/validate_graph_spec.py).
Audit Metadata