doc-coauthoring
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from various external platforms, creating a surface where malicious instructions could be embedded in the data being processed.
- Ingestion points: In
SKILL.md, the workflow describes reading content from user-provided files, templates, and integrations with services such as Slack, Microsoft Teams, Google Drive, and SharePoint. - Boundary markers: The instructions do not specify the use of delimiters or safety wrappers when processing this external context or when passing the document to sub-agents for 'Reader Testing'.
- Capability inventory: The skill uses file-creation (
create_file) and file-modification (str_replace) tools to draft and edit documents, and it has the capability to invoke sub-agents to analyze the generated content. - Sanitization: There is no evidence of content sanitization or instruction-filtering for the data retrieved from external integrations before it is used to draft the document or analyzed by sub-agents.
Audit Metadata