session-tracker

Track and recall session context across SeleneDB-backed agent sessions.

2|Updated Apr 3, 2026
One-click install
npx skills add https://github.com/jscott3201/ai-agent-skills --skill session-tracker-jscott3201
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: session-tracker
Source: https://github.com/jscott3201/ai-agent-skills/tree/main/skills/session-tracker
Command: npx skills add https://github.com/jscott3201/ai-agent-skills --skill session-tracker-jscott3201

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Maintain continuity across agent sessions by capturing rolling checkpoints and concise session summaries so context is not lost when a session ends or an agent restarts.

Core Features & Use Cases

  • Rolling checkpoints: Skills write session outcomes, summaries, and next steps to SeleneDB at completion so the graph reflects the latest state.
  • Automatic loading: SessionStart hook injects recent session context (recent summaries, active milestones, open findings) so new sessions start informed.
  • Agent coordination: Register agents, list peers, and provide heartbeat updates to avoid premature session reaping and to surface active collaborators.
  • Queries & views: Built-in history, search, and timeline queries let users and agents review recent work, find past decisions, and trace artifact chains.
  • Clean shutdown: Share session summary and deregister on end to release intents and provide short-term visibility.

Quick Start

Invoke the session-tracker skill with history, search <query>, or timeline to view recent session summaries and context.

Frequently Asked Questions about session-tracker

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I maintain agent session context across restarts?

To maintain agent session context across restarts, rolling checkpoints and concise summaries are written to SeleneDB at completion. This ensures prior context is captured and automatically injected into new sessions via a SessionStart hook.

What is the best way to track multi-agent collaboration history?

Tracking multi-agent collaboration history is handled through agent registration, heartbeat updates, and built-in timeline queries. This prevents premature session reaping and surfaces active peers alongside recent work summaries stored in SeleneDB.

Can I use SeleneDB to search past session summaries and decisions?

Yes, you can use SeleneDB to search past session summaries and decisions. The skill provides built-in history, search, and timeline queries that let users and agents review recent work and trace artifact chains directly.

How do agents load previous session context automatically?

Agents load previous session context automatically through a SessionStart hook that injects recent summaries, active milestones, and open findings. This ensures new agent sessions start informed without manual context retrieval.

Does session tracking work for debugging and project design sessions?

Session tracking works for debugging and project design sessions by applying rolling checkpoints to capture outcomes and next steps. This provides continuity for project work and reviews across multiple agent restarts and collaborations.

What happens to active sessions when an agent shuts down?

When an agent shuts down, a clean shutdown process shares the final session summary and deregisters the agent. This releases active intents, provides short-term visibility, and prevents the session from being prematurely reaped.