eval-dataset
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a workspace setup script (
scripts/setup_workspace.py) and multiple utility scripts (e.g.,normalize_traces.py,sample_traces.py,run_evals.py) intended for local automation and environment configuration. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data like support tickets and user feedback. It mitigates associated risks by prescribing boundary markers, such as keeping expected outputs in grader-only storage and redacting sensitive content at the collection boundary.
- [DYNAMIC_EXECUTION]: The installer script performs simple source-to-source adaptation by replacing workspace path placeholders in template files before writing them to the local filesystem.
- [EXTERNAL_DOWNLOADS]: The skill communicates with the author's official API (
api.confident-ai.com) for dataset management and instrumentation. These operations are consistent with the skill's purpose and utilize vendor-controlled infrastructure.
Audit Metadata