setup-project-skills
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands derived from the repository's own configuration files (e.g.,
package.jsonscripts,Makefiletargets). - Evidence: Step 5 (Verify) in
SKILL.mdexplicitly instructs the agent to "dry-run the cheapest configured command (e.g.commands.lintor--versionon the toolchain)" to validate the setup. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it parses various untrusted manifest and configuration files from the repository to influence its own behavior.
- Ingestion points:
package.json,pyproject.toml,Makefile,.github/workflows/, and other ecosystem-specific manifest files. - Boundary markers: The skill does not implement technical boundary markers for the data it reads, but it mitigates risk by requiring explicit user confirmation of all findings before proceeding.
- Capability inventory: The skill possesses file read/write capabilities and shell command execution via repository-defined runners.
- Sanitization: No explicit sanitization of the values extracted from repository manifests is described in the logic.
Audit Metadata