Manage Mentat audit events with JSONL validation and SQLite storage.
npx skills add https://github.com/matheushenriquefs/mentat --skill mentat-log
Please help me install this Agent Skill. Skill: mentat-log Source: https://github.com/matheushenriquefs/mentat/tree/main/.agents/skills/mentat-log Command: npx skills add https://github.com/matheushenriquefs/mentat --skill mentat-log
This Skill provides a comprehensive solution for managing and interacting with Mentat audit events, enabling structured logging, validation, and pruning of events for orchestration history and custom scripts.
Run the mentat-log skill to emit a new event:
mentat-log emit my-agent slice_started "Initial slice setup."
High-intent search queries and answers about installing and using this skill.
To emit structured audit events for Mentat orchestration, run the mentat-log skill with the emit command, specifying the agent name, event type, and message. This writes the event to the canonical SQLite store for orchestration history tracking.
Yes, you can validate JSONL log files for proper Mentat audit event format. The skill checks the structure of emitted events to ensure the integrity of audit logs for post-mortem analysis and custom script integration.
The best way to prune old agent log directories in Mentat is using the mentat-log skill's pruning feature. It removes outdated log directories based on specified criteria, maintaining a clean and organized audit trail.
Mentat audit logging does not require external dependencies. It utilizes Python's standard library for JSON handling and SQLite for database operations, storing structured events in a canonical local store without extra setup.
To list audit events from the canonical store in Mentat, use the mentat-log skill's listing feature. It retrieves and displays structured events from the SQLite database, allowing you to monitor orchestration history and external script interactions.
FindMalekConfigure Prisma for multiple database providers and set up the client.
imehrLog AI executions and learn from feedback using JSONL logs and JSON recommendations.
JansenAnalyticsStore entities and typed relations in a SQLite-backed graph with full-text search.
AcczdyImplements passwordless WebAuthn hardware key authentication with Python-fido2, Flask RP, and SQLite credential storage.