The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to manage project configuration and resolve repository paths. These commands (e.g., git rev-parse, mkdir, and ln) are scoped to the local filesystem and used for legitimate setup purposes as described in references/mapping.md.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface when processing user-provided transcription data. \n
Ingestion points: references/transcription.md (user-pasted meeting, course, or lecture content). \n
Boundary markers: Absent; the instructions do not implement specific delimiters or instructions to ignore embedded commands in the transcribed text. \n
Capability inventory: The agent has access to MCPVault tools (write_note, patch_note, search_notes) and local shell execution for configuration. \n
Sanitization: None specified; transcription content is processed directly to extract metadata, tags, and observations.

session-notes