obsidian
Warn
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of
obsidian-clivia a third-party Homebrew tap (yakitrak/yakitrak/obsidian-cli). This involves downloading and executing installation scripts and binaries from an external, non-verified source. - [COMMAND_EXECUTION]: The skill's primary functionality relies on executing the
obsidian-clibinary to perform file operations like search, create, move, and delete within the local file system. - [DATA_ACCESS]: The instructions direct the agent to read sensitive local application data at
~/Library/Application Support/obsidian/obsidian.jsonto identify vault paths and states. - [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the following surface:
- Ingestion points: The agent reads the content of arbitrary Markdown files (
*.md) within the Obsidian vault during search and edit operations. - Boundary markers: There are no instructions to use delimiters or ignore embedded instructions when processing note content.
- Capability inventory: The skill has the capability to write to the filesystem, rename files, and delete notes via
obsidian-cli. - Sanitization: There is no evidence of sanitization or validation for the content retrieved from notes before it is processed by the agent.
Audit Metadata