obsidian-cli
Warn
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill documents and encourages the use of the
obsidian evalcommand, which allows for the execution of arbitrary JavaScript code within the context of the Obsidian application instance. This functionality is primarily intended for developer workflows but represents a significant security surface if the input to the command is derived from untrusted sources. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a broad surface for processing untrusted data.
- Ingestion points: Vault content is ingested via
obsidian read,obsidian search, andobsidian daily:read. Developer logs and DOM content are ingested viaobsidian dev:consoleandobsidian dev:dom. - Boundary markers: The instructions do not define explicit boundary markers or delimiters for separating untrusted content from system instructions.
- Capability inventory: The skill has significant local capabilities including file deletion (
obsidian delete), file modification (obsidian create,obsidian append,obsidian move), and arbitrary code execution (obsidian eval). - Sanitization: There is no automated sanitization for the content read from the vault before it is processed by the agent.
- [COMMAND_EXECUTION]: The skill operates by executing the
obsidianCLI tool, which interfaces directly with the local file system and the running Obsidian application to perform vault mutations and configuration changes.
Audit Metadata