pr-to-video
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub Pull Requests, such as titles, descriptions, and diffs, which may contain malicious instructions.
- Ingestion points: Data is fetched via the GitHub CLI in scripts/fetch-pr.mjs and processed into a narrative source in scripts/ingest.mjs.
- Boundary markers: There are no explicit delimiters or instructions for the agent to ignore embedded commands within the PR data.
- Capability inventory: The skill can execute shell commands, write files, and perform network downloads.
- Sanitization: scripts/ingest.mjs strips HTML comments from the PR body to mitigate template-based injection.
- [EXTERNAL_DOWNLOADS]: Fetches contributor avatars from GitHub's official avatar hosting services.
- Evidence: scripts/fetch-people-avatars.mjs includes an SSRF guard restricting fetches to github.com and githubusercontent.com domains.
- [COMMAND_EXECUTION]: Orchestrates the workflow by executing local scripts and platform-specific CLI tools.
- Evidence: SKILL.md contains instructions to run scripts/fetch-pr.mjs, scripts/ingest.mjs, and npx hyperframes to ingest data and build the video.
Audit Metadata