Chron
Keep a local, timestamped audit trail of AI conversations in SQLite so you can review what the agent said and did without sending logs to the cloud.
Overview
Chron is a MCP server for the Operate phase that stores timestamped AI conversation audit logs locally in SQLite.
What is this MCP server?
- Writes timestamped audit records for AI conversations into local SQLite
- You own the database file—no vendor cloud log lock-in
- CHRON_DB_PATH configures where the SQLite store lives
- npm chron-mcp 0.1.7 with stdio MCP transport
- Fits privacy-conscious solo builders who need a paper trail of agent sessions
- Server version 0.1.7 as npm identifier chron-mcp
- One documented env var: CHRON_DB_PATH
What problem does it solve?
When agent chats vanish between sessions, you cannot prove what was promised, debug regressions, or review risky advice after the fact.
Who is it for?
Privacy-minded solo builders who want a local conversation paper trail for support, compliance, or post-incident review.
Skip if: Teams that need centralized multi-user analytics, real-time dashboards, or cloud-hosted log search at scale.
What do I get? / Deliverables
After you register Chron, each conversation can be persisted to your SQLite file so you can audit and revisit agent activity on your own terms.
- Persistent timestamped conversation audit records in SQLite
- Configurable local database location via CHRON_DB_PATH
- Stdio MCP integration for ongoing session logging
Recommended MCP Servers
Journey fit
Running agent workflows in production-like rhythms needs observability over prompts, replies, and decisions you can replay later. Conversation audit logging is operational monitoring: it helps you trace incidents, compliance questions, and support escalations on your own machine.
How it compares
Local conversation audit MCP, not a full observability platform or error-tracking service.
Common Questions / FAQ
Who is Chron for?
Builders and operators who run AI agents daily and want an owned, timestamped history of what was discussed without cloud logging.
When should I use Chron?
Use it once you depend on agents for production-adjacent work and need to monitor, review, or audit past conversations from your machine.
How do I add Chron to my agent?
Register the chron-mcp npm package as a stdio MCP server in your agent config and set CHRON_DB_PATH to your preferred SQLite file path.