table-api-reads

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill describes processes for reading and writing data to external ServiceNow records, which may contain untrusted content.
  • Ingestion points: Data is ingested through tools described in SKILL.md, including snow_query_table, snow_get_by_sysid, and snow_record_manage.
  • Boundary markers: The instructions do not specify the use of delimiters or markers to separate untrusted record content from agent instructions.
  • Capability inventory: The skill identifies tools for querying data, adding comments (snow_add_comment), and pulling code artifacts (snow_pull_artifact).
  • Sanitization: The documentation does not mention sanitizing or escaping content from external records before processing.
  • [COMMAND_EXECUTION]: The documentation describes the side effects of the snow_pull_artifact tool, which involves filesystem operations.
  • Evidence: As noted in SKILL.md, the tool writes files to <tmpdir>/serac-artifacts/<table>/<name> on the local filesystem when using the stdio transport mechanism.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 07:16 PM
Security Audit — agent-trust-hub — table-api-reads