stitch-workflow
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute shell commands to download generated design assets and initialize development environments.
- Evidence: Use of
curl -sL {downloadUrl}to fetch HTML source code from dynamic URLs. - Evidence: Use of
npx storybook@latest initto scaffold Storybook within the project. - [EXTERNAL_DOWNLOADS]: Fetches design files and HTML markup from external URLs provided by MCP tools.
- Evidence: The agent is instructed to download standalone Tailwind HTML documents via
curl. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local configuration files and remote downloads, creating a vector for instructions to be embedded in data.
- Ingestion points: Reads session settings from
.codex/stitch.local.mdor.claude/stitch.local.md, and content from external download URLs. - Boundary markers: None present for distinguishing between data and instructions in the ingested files.
- Capability inventory: Capabilities include file system writes, network requests, and execution of various specialized designer tools.
- Sanitization: No sanitization or validation of the ingested content is described before the agent processes it for component extraction.
Audit Metadata