anki
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill configuration relies on fetching the
anki-mcp-servertool from the npm registry using thenpxutility. - [COMMAND_EXECUTION]: The agent executes the
anki-mcp-serverprocess locally to translate agent tool calls into API requests for the Anki application. - [INDIRECT_PROMPT_INJECTION]: The skill acts as an ingestion surface for external data by processing user-provided text and knowledge points into flashcard content.
- Ingestion points: Textual content provided by the user for extraction into flashcards as described in
SKILL.mdworkflows. - Boundary markers: None explicitly defined in the prompt instructions to delineate untrusted data.
- Capability inventory: The skill allows creating, updating, searching, and deleting notes within the local Anki database.
- Sanitization: Validation and sanitization are handled by the underlying MCP server implementation and the AnkiConnect addon interface.
Audit Metadata