spec-to-repo
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes arbitrary natural-language specifications to generate executable code, creating a surface for malicious instructions to influence the generated application logic.
- Ingestion points: Project specifications provided by the user in SKILL.md and interpreted via references/spec-parsing-guide.md.
- Boundary markers: Absent. The skill lacks instructions for the agent to distinguish between legitimate project requirements and embedded adversarial instructions within the user's spec.
- Capability inventory: The skill can write files, create directories, and execute the local validation script scripts/validate_project.py.
- Sanitization: The included scripts/validate_project.py focuses on best practices (missing README, committed .env files, placeholders) but does not analyze the generated code for malicious behavior, logic bombs, or exfiltration endpoints.
- [COMMAND_EXECUTION]: The skill executes a local Python script, scripts/validate_project.py, to perform structural and best-practice checks on the generated project files. This is an intended functionality for quality assurance.
Audit Metadata