trellis-spec-bootstarp
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes repository source code and configuration files which could contain malicious instructions. Ingestion points: Repository source code, manifests, and workspace configs (references/repository-analysis.md). Boundary markers: None identified. Capability inventory: Shell command execution and file operations (references/mcp-setup.md, references/spec-writing.md). Sanitization: None identified.
- [EXTERNAL_DOWNLOADS]: The skill fetches tools from external sources, including GitNexus via NPM and the ABCoder tool from the CloudWeGo repository on GitHub.
- [REMOTE_CODE_EXECUTION]: The setup process executes code downloaded from external registries, specifically using 'go install' for ABCoder and 'npx' for GitNexus.
- [COMMAND_EXECUTION]: The skill workflow involves executing shell commands to perform analysis and verify tool installations, such as 'npx gitnexus status' and 'abcoder parse'.
Audit Metadata