todoist
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes Todoist MCP tools including
find-tasks,complete-tasks, andreschedule-tasksto manage the lifecycle of user tasks. This is intentional behavior for the skill's purpose. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from Todoist task titles and descriptions to categorize items and generate Obsidian notes. 1. Ingestion points: Todoist task titles and descriptions fetched via
find-tasks. 2. Boundary markers: No explicit delimiters or instructions are used to separate task content from agent directives. 3. Capability inventory: Modification of Todoist tasks (completing/rescheduling) and writing to a local filesystem (Obsidian vault). 4. Sanitization: No sanitization or escaping of task content is performed before interpolation into the Obsidian capture target.
Audit Metadata