brightdata-debug-bundle
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses Bash to execute connectivity tests (curl, nc) and filesystem operations (mkdir, tar) to generate a diagnostic bundle.
- [DYNAMIC_EXECUTION]: The skill dynamically generates a shell script (brightdata-debug-bundle.sh) to consolidate diagnostic logic and package results.
- [INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection by ingesting untrusted data from local log files during the debug collection process.
- Ingestion points: Reads all files matching logs/*.log (Step 5 in SKILL.md).
- Boundary markers: None implemented for the content of the logs being aggregated.
- Capability inventory: Uses Bash, curl, and tar (Step 3, 4, 5 in SKILL.md).
- Sanitization: Implements basic password redaction in log files using sed and provides a manual checklist for user review before sharing the bundle.
Audit Metadata