reconciliation
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (authorized report pages) which could contain instructions intended to influence the agent's behavior.
- Ingestion points: Tools listed in
references/sandbase-api-map.md(context_dev_scrape_markdown,context_dev_extract_structured_data) are used to pull content from external report pages into the agent's context. - Boundary markers: The skill includes explicit warnings in
SKILL.mdto work from copies rather than source records and mandates that all outputs be reviewed by qualified financial professionals. - Capability inventory: The skill is limited to data extraction and comparison for the purpose of identifying reconciling items. Instructions in both
SKILL.mdandreferences/sandbase-api-map.mdexplicitly prohibit modifying source accounting records or posting journal entries. - Sanitization: The skill relies on a human-in-the-loop review process as the primary control for the generated reconciliation artifacts.
Audit Metadata