workers-best-practices
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches documentation from developers.cloudflare.com and downloads the @cloudflare/workers-types package from the npm registry using npm pack to provide the agent with current API definitions. These sources belong to a trusted organization and are documented neutrally.
- [COMMAND_EXECUTION]: Employs shell commands including npm pack, tar, cat, and mkdir to retrieve type information and read project configuration files like wrangler.jsonc during the review process.
- [INDIRECT_PROMPT_INJECTION]: The skill has an inherent attack surface for indirect prompt injection by design, as it must ingest and process untrusted user-supplied source code and configuration files.
- Ingestion points: project source files (.ts, .js) and wrangler.jsonc processed during the review workflow.
- Boundary markers: The skill instructions do not specify delimiting markers or ignore-rules for the content being analyzed.
- Capability inventory: The skill utilizes shell command execution (npm, tar, cat) and network requests to Cloudflare documentation.
- Sanitization: No sanitization is performed on the code content before it is processed by the agent.
Audit Metadata