expo-router
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent reports errors to the vendor using content derived from user-provided files, creating a potential surface for instruction injection.\n
- Ingestion points: User-provided source code, project directory structures, and error logs analyzed by the agent (SKILL.md).\n
- Boundary markers: The instructions do not specify any delimiters or warnings to prevent the agent from obeying instructions embedded within the feedback text (SKILL.md).\n
- Capability inventory: The agent is instructed to execute shell commands using the
npxutility (SKILL.md).\n - Sanitization: No sanitization or escaping mechanisms are described for the
<actionable feedback>placeholder in the command string (SKILL.md).\n- [COMMAND_EXECUTION]: The skill documentation includes a pre-formatted shell command intended for the agent to use when reporting issues.\n - Evidence:
npx --yes submit-expo-feedback@latest --category skills --subject "expo-router" "<actionable feedback>"(SKILL.md).\n- [EXTERNAL_DOWNLOADS]: The feedback mechanism fetches and executes the latest version of a reporting tool from the NPM registry at runtime.\n - Evidence: The use of
npx --yes submit-expo-feedback@latestinvolves downloading external code from a well-known package registry (SKILL.md).
Audit Metadata