heyeddi-setup
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to manage project setup preferences in
.heyeddi/stack.json. All filesystem operations are restricted to the project root and sibling skill directories for the purpose of metadata indexing. - [INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface where user-provided setup 'answers' are written to a configuration file that the agent later reads.
- Ingestion points:
scripts/write_setup.pyvia the--jsonargument. - Boundary markers: None; values are stored in a structured JSON format.
- Capability inventory: File read/write capabilities (
pathlib.Path.write_text,pathlib.Path.read_text). - Sanitization: The skill implements a schema validation layer in
scripts/_stack_schema.py(validate_value,normalize_custom_workflow) which restricts inputs to specific types (strings, booleans) and allowed values for key fields (presets, agent autonomy settings). Given these constraints, the risk of instruction injection via the configuration file is negligible.
Audit Metadata