soia-dev-draft-test-doc
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation provides an installation command for a package hosted in the vendor's repository (
soia-team/soia-open-dev-testing-skills). This is a standard deployment mechanism for the skill and matches the author's provenance. - [DATA_EXFILTRATION]: The skill includes explicit instructions to avoid reading unauthorized data, credentials, or private repositories. It specifies that any temporary data should be stored in system temporary directories and cleared after use, minimizing the risk of data exposure.
- [PROMPT_INJECTION]: The skill processes user-supplied requirements, which inherently involves an indirect prompt injection surface. However, the skill's logic is constrained to generating Markdown documentation and does not utilize dangerous tools or shell commands that could be exploited via injection. It also specifies the use of anonymized or placeholder data when processing sensitive requirements.
- Ingestion points: User-provided requirement texts, PRDs, and prototype descriptions in SKILL.md.
- Boundary markers: Absent; the skill uses internal processing workflows and structured output templates rather than explicit boundary delimiters.
- Capability inventory: None; the skill is restricted to generating text/Markdown output within the agent session.
- Sanitization: Absent; the instructions do not explicitly describe input sanitization techniques.
Audit Metadata