baoyu-format-markdown
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided markdown or plain text files to analyze and format them. Malicious content within these input files could attempt to influence the agent's behavior during the analysis (Step 2) or formatting (Step 4) phases.
- Ingestion points: The user-specified file is read into the agent's context in Step 1 of the workflow.
- Boundary markers: Absent. The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when reading the external file content.
- Capability inventory: The skill has the capability to write to the file system (analysis and formatted files) and execute local scripts (
scripts/main.ts) viabunornpx. - Sanitization: The typography script uses AST-based parsing via the
unifiedandremarkecosystems, which provides a structured and relatively safe way to handle markdown content. - [EXTERNAL_DOWNLOADS]: The typography script executes an external utility package at runtime using
npx. - Evidence: The file
scripts/autocorrect.tsusesspawnSyncto runnpx autocorrect-node --fix, which fetches theautocorrect-nodepackage from the npm registry if it is not already available locally.
Audit Metadata