convex-docs
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from external documentation sites and local package files, creating a surface for indirect prompt injection if that data contains malicious instructions.
- Ingestion points: Fetches content from
docs.convex.devand reads type definitions/metadata fromnode_modulesas described inSKILL.md. - Boundary markers: Absent; the instructions do not specify the use of delimiters or 'ignore' instructions when the agent processes the fetched documentation content.
- Capability inventory: The skill utilizes shell command execution (
node -p), file system read access (package.json,node_modules), and network operations (docs.convex.dev). - Sanitization: Absent; there is no mention of sanitizing or escaping the content retrieved from external documentation pages.
- [COMMAND_EXECUTION]: The workflow utilizes a shell command to identify the version of the installed software library.
- Evidence:
node -p "require('./node_modules/convex/package.json').version"inSKILL.md. This command is used locally to ensure documentation matches the project environment, which is a common development practice.
Audit Metadata