build-store
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes
npx foundax@latestto retrieve and run the author's command-line interface tool directly from the npm registry. - [REMOTE_CODE_EXECUTION]: The workflow involves executing the author's official code locally to initialize projects, run a local preview server, and package the store for deployment.
- [INDIRECT_PROMPT_INJECTION]: The instructions direct the agent to read and follow a workflow defined in
AGENTS.md, a file created during the project initialization step. This represents an attack surface where instructions inside project files influence the agent's behavior. - Ingestion points:
AGENTS.md(read in step 3) - Boundary markers: Absent; the agent is told to follow the workflow from the first step without specific delimiters.
- Capability inventory: Shell command execution via
npx(init,check,preview,pack). - Sanitization: Absent; there is no explicit validation of the content within
AGENTS.mdbefore the agent follows its instructions.
Audit Metadata