Skills
Navigate
skills/bruin-data/bruin/add-ingestr-source/Gen Agent Trust Hub

add-ingestr-source

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process information from external sources, including documentation, PRs, and the ingestr repository's source code. This creates a surface where instructions embedded in those external materials could influence the agent's behavior.
  • Ingestion points: Upstream documentation (getbruin.com), third-party PRs, and the bruin-data/ingestr repository code.
  • Boundary markers: No explicit boundary markers or warnings to ignore embedded instructions are provided.
  • Capability inventory: The skill has capabilities to modify Go code, update YAML/JSON configuration files, and execute shell commands (go run, make).
  • Sanitization: No specific sanitization or validation of the external content is mentioned.
  • [COMMAND_EXECUTION]: The skill workflow involves executing local shell commands to regenerate schema expectations and run project tests.
  • Evidence: The skill instructs running go run -tags="no_duckdb_arrow" . internal connections | jq . > integration-tests/expectations/expected_connections_schema.json to update the connection schema, as well as make format and make test for validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 11:18 AM