report-automation
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting data from external systems (APIs and unstructured sources) and processing that data through an LLM for narrative generation. This creates an attack surface for indirect prompt injection where malicious content within the source data could influence the report's narrative output.\n
- Ingestion points: External data extracted from Stripe, PostHog, and GitHub APIs, as well as unspecified unstructured sources referenced in SKILL.md.\n
- Boundary markers: The instructions do not define specific delimiters or guardrails to prevent the LLM from obeying instructions embedded in the ingested data during the narrative synthesis phase.\n
- Capability inventory: The skill framework leverages LLM synthesis for report generation; no direct script-based command execution or file-write capabilities were found in the provided files.\n
- Sanitization: There are no requirements specified for the filtering or sanitization of ingested content before it is passed to the LLM for summarization.
Audit Metadata