obsidian-incident-runbook
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to search through and process the contents of markdown files within the user's vault to detect corruption, sync conflicts, and broken metadata. If these user-controlled files contain adversarial instructions, they could potentially influence the agent's behavior during the diagnostic process.
- Ingestion points: SKILL.md (Step 3: Check for corrupted files; Step 4: Sync Conflict Resolution).
- Boundary markers: Absent. The skill uses standard grep and head commands without delimiters or warnings against embedded instructions.
- Capability inventory: The skill uses Read, Grep, and Bash tools to examine and modify files.
- Sanitization: Absent. Content from the markdown files is processed directly via shell utilities.
- [DYNAMIC_EXECUTION]: The skill employs inline Python scripts to programmatically read and modify Obsidian's JSON configuration files, such as community-plugins.json and appearance.json. These scripts are used for administrative tasks like binary searching for buggy plugins or resetting UI themes.
Audit Metadata