skills/kensaurus/skills/docs-coauthor/Gen Agent Trust Hub

docs-coauthor

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to gather context from various external and potentially untrusted sources (e.g., Slack threads, Teams messages, Google Drive, SharePoint) via integrations or direct user input. This gathered information is then used to inform the drafting process and is passed to sub-agents during the 'Reader Testing' phase, creating an opportunity for embedded malicious instructions to influence the agent's behavior.
  • Ingestion points: The skill (in Stage 1: Context Gathering) explicitly instructs the agent to read shared documents, fetch content from links, and utilize integrations for messaging apps (Slack, Teams) and storage services (Google Drive, SharePoint).
  • Boundary markers: The skill lacks defined delimiters or system instructions that would help the agent distinguish between informational content and potential commands embedded within the external data.
  • Capability inventory: The skill possesses the capability to read from external integrations, perform in-place edits on local documents, and programmatically invoke sub-agents to process the document content.
  • Sanitization: There is no evidence of filtering, escaping, or validation of the external content before it is interpolated into prompts or shared with sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 11:26 PM