living-one-pager
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data from user conversations, session histories, and external documents to synthesize product accounts.
- Ingestion points: Agent session history, external links, and user-supplied documents as described in
SKILL.mdandreferences/workflow.md. - Boundary markers: Instructions direct the agent to distinguish between source roles and assistant summaries, preserving original context.
- Capability inventory: Writing to local files (
./work-items/), Gists, or wikis, and reading from session history and provided files. - Sanitization: The skill includes a dedicated reference file (
references/public-export.md) providing detailed procedures for sanitizing sensitive information (names, emails, credentials, internal paths) before public sharing.
Audit Metadata