- Home
- Skills
- Agents & Orchestration
- officelapse
Works with the AI Agents you already use
officelapse
Visualize Claude Code sessions and subagents as an interactive, 3D isometric office with live replay and rewind.
Free
officelapse
Example session with this skill installed
I want to see what my agents have been doing today. Can you visualize the sessions?
- Read your context and instructions
- Compiled the officelapse
I have prepared the visualization environment. I checked your Python version, updated the local officelapse repository, and started the visualizer for the last 24 hours of activity.
Visualization Server Active
The 3D isometric office is now running. You can see your subagents at their desks and navigate the timeline to see previous actions.
- URL: http://localhost:8777
- Source Logs:
~/.claude/projects - Timeframe: Last 24 hours
Controls:
- Pan/Zoom: Drag to move, scroll to zoom in on specific rooms.
- Agent Detail: Hover over a teammate to see their current task or log output.
- Enter Room: Click a room to step inside for a focused view.
- Timeline: Drag the slider at the bottom to rewind to an earlier state.
- Search: Press
/to search for specific sessions or agent names.
Next steps
- Open the URL in a WebGL-compatible browser to view the office.
- To stop the server, run
kill $(lsof -ti :8777). - To view a longer history, relaunch with
HOURS=72 ./run.sh.
Connects securely to your tools. The creator never sees your data.
About this skill
Visualizing agent activity shouldn't require digging through raw log files or terminal scrollback. This skill turns Claude Code session data into a live, interactive 3D isometric office.
What it does
- Live visualization showing Claude Code subagents as animated teammates at desks.
- Session playback allowing you to rewind and fast-forward through previous agent actions using a timeline.
- Deep dive inspection providing detailed context when hovering over specific agents or clicking into team rooms.
- Search and navigation including spatial controls for panning, zooming, and searching session history via keyboard shortcuts.
- Local execution ensuring all session data remains on your machine by reading local log files directly.
How it works
- Environment check verifying Python 3.8+ availability on the local system.
- Repository management cloning the tool or updating an existing installation from GitHub.
- Instance launch executing the provided shell script with optional port and timeframe parameters.
- Interactive viewing guiding you to the local web interface to monitor or replay sessions.
Frameworks & tools
The tool runs as a local Python-based web app. It requires a browser with WebGL support to render the 3D environment and reads data directly from the ~/.claude/projects directory.
Why this beats prompting it yourself
Parsing nested Claude Code session logs manually is time-consuming and fails to show the parallel nature of subagent work. This skill provides an immediate, spatial representation of agent orchestration that a text-based prompt cannot replicate.
Use cases
- Debugging complex multi-agent workflows by replaying the sequence of events.
- Monitoring long-running agent tasks in a peripheral dashboard.
- Reviewing "what my agents did today" across multiple project directories.
Known limitations
Supports Claude Code sessions only. Does not visualize Cursor, Copilot, Aider, or other IDE agents. Requires a WebGL-compatible browser and sessions within the specified timeframe.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Only visualizes Claude Code sessions. Requires Python 3.8+ and a WebGL-enabled browser. Older logs are hidden unless the HOURS variable is adjusted.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
2 uses
Times used to date
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 12 days ago
- Free with an Agensi account
- Software
Git
Python 3.8+A browser with WebGL