The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface detected within the vault-to-skill synchronization process. Ingestion points: Data is ingested from external Markdown files located in the 'Second Brain' directory using search and read operations. Boundary markers: No delimiters or 'ignore' instructions are used to isolate ingested pattern text from the agent's primary instructions. Capability inventory: The skill can modify other SKILL.md files and executes obsidian CLI commands to interact with the filesystem. Sanitization: No validation or sanitization of the proposed pattern content is performed before presentation for approval.
[COMMAND_EXECUTION]: The skill executes local commands using the obsidian CLI tool to search, read, and append data within the local vault and updates agent behavior by modifying SKILL.md instruction files.

knowledge-sync