pasta-objectives
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains no executable code, scripts, or network operations. It consists entirely of markdown instructions for a methodology workflow.
- [DATA_EXPOSURE]: The workflow involves reading project configuration files such as
.env.example,package.json, and database migrations. This behavior is documented as the primary function of the skill (threat modeling context gathering) and does not involve exfiltrating data to external sources. - [PROMPT_INJECTION]: The instructions do not contain any patterns typical of prompt injection, such as attempts to bypass safety filters, override system instructions, or extract system prompts.
- [INDIRECT_PROMPT_INJECTION]: The skill defines an ingestion surface by instructing the agent to parse untrusted user content (e.g., README files, API routes, OpenAPI specs). This is a known risk for indirect prompt injection where malicious instructions could be embedded in the codebase being analyzed. However, the skill does not possess dangerous capabilities like shell execution or network requests that would allow such an injection to be weaponized locally.
- [EXTERNAL_DOWNLOADS]: No external downloads or remote script execution patterns were identified. All referenced paths are relative internal documentation links.
Audit Metadata