outgoing-link-audit
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local repository script (
scripts/check_external_links.sh) and standard development commands includingnpm run build,npm test, andgit rev-parse. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external websites to audit their status, creating a potential surface for indirect prompt injection.
- Ingestion points: Untrusted content is ingested when the agent fetches external URLs extracted from local build files (
dist/*.html). - Boundary markers: No specific boundary markers or instructions are used to separate fetched web content from agent instructions.
- Capability inventory: The agent has the capability to execute shell scripts, perform network requests, and modify repository source files (
src/pages/*.astro) and test suites (tests/links.spec.ts). - Sanitization: Content retrieved from external URLs is not sanitized or filtered before the agent is asked to "apply judgment" to the results.
- [DATA_EXFILTRATION]: The skill performs automated network requests to various external third-party domains discovered within the project's link tags to verify their availability and redirection status.
Audit Metadata