dcoon-notebook
Pass
Audited by Gen Agent Trust Hub on Nov 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions explicitly direct the agent to read a file named
AGENTS.mdwithin the notebook to understand how to perform its tasks. This creates a vulnerability where malicious instructions embedded in that file could influence the agent's behavior. - Ingestion points: The agent is instructed in
SKILL.mdto read<notebook>/AGENTS.mdimmediately after opening any notebook or wiki to determine reading and writing procedures. - Boundary markers: There are no specified delimiters or instructions provided to the agent to treat the content of
AGENTS.mdas data rather than instructions. - Capability inventory: The agent is granted capabilities to search, read, patch, and write files within the vault using MCP tools, CLI tools, or direct file system access.
- Sanitization: No sanitization or validation of the content within
AGENTS.mdis required by the skill before the agent processes it.
Audit Metadata