playwright-trace-analyzer
Pass
Audited by Gen Agent Trust Hub on May 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads test artifacts (trace.zip) from GitHub Actions using the
ghCLI. This is a core feature for analyzing CI failures from a well-known service. - [COMMAND_EXECUTION]: The skill executes several local Node.js scripts (
fetch-gh-run.mjs,trace-extract.mjs, etc.) and system utilities includinggh,unzip, andjqto process trace data. The scripts include input sanitization (regex-based) to prevent path traversal when handling artifact names. - [PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection because it processes and displays verbatim content from external traces, such as console logs and page errors.
- Ingestion points:
trace.traceandtrace.network(JSONL format) processed viascripts/trace-extract.mjs. - Boundary markers: Absent; the instructions tell the agent to quote error messages and logs verbatim without providing delimiters or warnings to ignore potential instructions in the data.
- Capability inventory: The agent can execute system commands (
gh,jq,unzip) and Node.js scripts which can perform file system operations. - Sanitization: Content is parsed as JSON but the resulting textual data is not filtered or escaped before presentation to the agent.
Audit Metadata