A proactive state management skill that ensures AI agents maintain persistent task awareness across session resets and background processes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install sunshine-agent-task-tracker
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install sunshine-agent-task-tracker using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Task Tracker?
Task Tracker is a vital addition to your collection of Openclaw Skills designed to solve the problem of context loss during session resets or memory compaction. By maintaining a live state file, this skill acts as a bridge between sessions, allowing an AI agent to instantly resume work with full knowledge of current, background, and pending tasks.
This skill functions as a single source of truth for the agent's operations. It proactively logs every state change—from initial request to final result—ensuring that even if a session is interrupted, the agent can read the state file and immediately understand what was requested, what processes are still running in the background, and what steps remain to be completed.
Task Tracker Use Cases
- Resuming complex development tasks after an AI session reset or context window compaction.
- Monitoring long-running background processes, SSH sessions, and PIDs across different servers.
- Maintaining a concise history of recent accomplishments without bloating the primary conversation context.
- Ensuring multi-step workflows remain on track by recording progress details after every individual step.
How Task Tracker Works
- The agent receives a task and immediately initializes an entry in memory/tasks.md with a status of in-progress.
- Any initiated background processes are recorded with specific metadata, including session IDs, PIDs, and server names.
- The agent updates the state file before every user report to ensure the physical log reflects the internal logic (write-first protocol).
- Upon task completion or failure, the entry is moved to the recent history section with a final summary or error log.
- At the start of every new session, the agent reads the state file to re-synchronize its awareness with the project's current status.
Task Tracker Setup
To integrate this into your environment, ensure your agent has read/write access to a local memory directory. You can initialize the state file using the following structure:
mkdir -p memory
touch memory/tasks.md
No manual user invocation is required; once configured, Openclaw Skills like Task Tracker trigger automatically based on agent activity.
Task Tracker Data Schema & Taxonomy
The skill manages data within memory/tasks.md using a strict, lightweight Markdown schema to keep the file under 2KB.
| Field | Description |
|---|---|
| Status | Visual indicator: 🔄 In-progress, ✅ Complete, ❌ Failed, or ⏸️ Paused. |
| Background | Technical details for remote tasks: session-id, PID, and server. |
| Notes | Concise progress details and partial results. |
| Pruning | Completed tasks older than 3 days are automatically removed to save tokens. |
Task Tracker Advanced Features
- Automatic background process tracking for persistent SSH and CLI commands.
- Session-resilient awareness that survives context window refreshes.
- Write-first rule ensuring state is saved before communication to prevent data loss.
- Automated pruning logic to maintain a sub-50 line file for optimal token efficiency.
- Cross-session memory synchronization using Openclaw Skills architecture.
SKILL.md
Loading
Related Openclaw Skills
Super RSS Agent
A powerful RSS management and reading tool for Openclaw Skills that automates subscription tracking, article summarization, and content discovery.
Bybit V5 OpenAPI Skill
A specialized interface for querying Bybit V5 public market data using uxc and curated OpenAPI schemas.
Automation Workflows
A strategic framework for solopreneurs to identify, design, and implement high-ROI no-code automation workflows that reclaim valuable time.
Skill Creator
The Skill Creator is a comprehensive framework designed to help developers build, test, and refine high-performance AI agent skills through data-driven iteration and benchmarking.
Trigger
A versatile command-line event trigger toolkit designed for managing webhooks, file watchers, and condition-based automation.
Haodf Doctor Search
A medical navigation skill that connects users with the right healthcare professionals through symptom analysis and specialty mapping.
Featured*
