- Home
- Skills
- Agents & Orchestration
- tidewatch
Works with the AI Agents you already use
tidewatch
A live status dashboard for managing and monitoring multiple concurrent AI agent sessions from a single view.
$9.99
tidewatch
Example session with this skill installed
Set up a new Tidewatch session for the 'api-refactor' project to track the migration to FastAPI. We are starting with the Auth module.
- Read your context and instructions
- Compiled the tidewatch
I have initialized the Tidewatch session and generated the live dashboard. You can monitor the progress of this refactor at ~/.tidewatch/tidewatch.html.
Session Initialized: api-refactor-claude-x7b2
{
"agent": "claude",
"project": "api-refactor",
"title": "FastAPI Migration",
"task": "Auth Module Migration",
"goal": "Migrate existing OAuth2 logic to FastAPI dependencies",
"status": "working",
"now": "Analyzing current auth middleware structure",
"needsYou": "",
"steps": [
{ "text": "Map current auth routes", "state": "doing" },
{ "text": "Define FastAPI security dependencies", "state": "todo" },
{ "text": "Implement JWT validation logic", "state": "todo" },
{ "text": "Verify with integration tests", "state": "todo" }
],
"log": ["Session started for api-refactor"]
}
Next steps
- Open the dashboard using
node ~/.tidewatch/tidewatch.mjs appto see the live progress. - I will update the board as soon as the mapping of auth routes is complete.
- Check for a desktop notification if I encounter a dependency conflict.
Connects securely to your tools. The creator never sees your data.
About this skill
Context switching between a dozen AI agent sessions leads to mental fatigue and lost progress. Tidewatch solves this by providing a unified live dashboard where every active session reports its status, roadmap, and blocks in a single view.
What it does
- Session tracking maintains individual JSON states for multiple concurrent agents, including current task, goal, and log.
- Visual roadmaps renders a step-by-step progress board showing completed, active, and errored stages.
- Status monitoring updates session health as working, waiting, blocked, or done with associated visual states.
- OS notifications triggers system-level alerts for macOS, Windows, and Linux when an agent completes a task or hits a blocker.
- Unified dashboard builds a lightweight HTML frontend that auto-refreshes every three seconds without requiring a backend server.
How it works
- Initialize session by generating a unique ID and creating a state JSON in the
~/.tidewatch/sessions/directory. - Update state by rewriting the JSON file whenever a significant milestone, error, or context shift occurs.
- Build dashboard by running the build command to compile all session data into a single HTML view and data manifest.
- Monitor pets optionally launches a desktop pet on Windows that tracks overall progress on the taskbar.
Frameworks & tools
The system uses Node.js for state management and build scripts. It operates locally using the user home directory for signal persistence and generates a standalone vanilla JavaScript and HTML dashboard.
Why this beats prompting it yourself
Manual status reporting is inconsistent and quickly becomes stale as the agent iterates. This skill enforces a structured schema and automated build pipeline that ensures your dashboard reflects reality across 10+ terminal windows without manual intervention.
Use cases
- Manage parallel microservice refactors across multiple Claude Code instances.
- Track long-running background tasks while focusing on a separate terminal session.
- Identify at a glance which agent is stuck on a permission prompt or test failure.
Known limitations
The system requires Node.js installed locally. Writing to the home directory might require sandbox adjustments for certain agent environments.
Requires Node.js runtime. Agents running in strict sandboxes may need explicit directory write permissions for ~/.tidewatch. Multi-machine syncing is not supported.
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 any payment.
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
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 7 days ago
Requires
Permissions
File Scopes
Network access is local only: the board is served from 127.0.0.1 with a per-run token. No data leaves your machine. ~/.claude/settings.json is changed only if you choose to connect usage (status line); the original is backed up first. Claude/Codex session logs are read only for model and usage info, never message content.