A proactive state management system that ensures AI agents maintain progress context across session resets and complex multi-step workflows.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install 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 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?
The Task Tracker is a specialized utility designed to bridge the context gap during AI agent session resets or memory compaction events. By maintaining a single source of truth in a localized markdown file, it ensures that every action, background process, and partial result is recorded immediately. This skill is a core component for developers looking to build robust workflows within the Openclaw Skills ecosystem, providing a reliable way for agents to keep track of their current objectives without relying solely on conversation history.
It functions as an autonomous state engine, updating its records before communicating with the user. This write-first approach guarantees that even if a session crashes or needs to be restarted, the agent can resume awareness by reading the stored state. This makes it particularly effective for long-running background tasks, SSH sessions, and multi-stage deployments where continuity is critical for productivity.
Task Tracker Use Cases
- Maintaining context during AI agent session resets or memory compaction.
- Tracking long-running background processes including PIDs and server details.
- Resuming multi-step development workflows after a break or interruption.
- Providing a consolidated source of truth for active and recently completed tasks.
- Coordinating between multiple agents or tools by sharing a synchronized state file.
How Task Tracker Works
- The agent detects a new task request and immediately initializes an entry in the state file with an active status.
- Upon starting background processes, it records technical metadata such as session IDs, PIDs, and specific commands to ensure traceability.
- As progress occurs, the agent updates the status and notes fields before reporting back to the user.
- On task completion or failure, the entry is updated with final results or error logs and moved to the recent history section.
- At the start of every new session, the agent reads the state file to re-establish environmental awareness and resume pending activities.
Task Tracker Setup
To integrate this into your environment, create the necessary directory structure and ensure the agent has read/write permissions for the memory directory.
mkdir -p memory
touch memory/tasks.md
Configure your agent to trigger the Task Tracker update logic on every significant state change. Ensure the agent is instructed to read the file during its initialization phase to leverage the full power of Openclaw Skills.
Task Tracker Data Schema & Taxonomy
The Task Tracker organizes data in a compact Markdown format within memory/tasks.md to ensure it fits within agent context limits.
| Field | Description | Format/Example |
|---|---|---|
| Status | Current state of the task | 진행중 (In Progress), 완료 (Completed), 실패 (Failed) |
| Requested | Timestamp of the initial request | YYYY-MM-DD HH:MM |
| Background | Metadata for remote or async jobs | session-id (PID) on server-name |
| Notes | Incremental progress details | Partial results and step-by-step updates |
| Result | Final output or summary | Reference links and closing data |
Note: The file is restricted to 50 lines or 2KB to maintain high-speed processing.
Task Tracker Advanced Features
- Automatic pruning of completed tasks older than three days to maintain high performance.
- Write-first protocol ensures data integrity even during unexpected agent crashes or environment resets.
- Support for tracking persistent SSH sessions and background PID management across different servers.
- One-line summaries for completed tasks to maximize context window efficiency during long-term projects.
- Explicit Empty Active indicators to prevent agent hallucination regarding current workloads when no tasks are pending.
SKILL.md
Loading
Related Openclaw Skills
Salesforce Query (Schema-Adaptive)
A schema-adaptive Salesforce integration for Openclaw Skills that discovers environment-specific data structures to enable resilient GTM workflows.
Immersive AR/VR Marketing for High-Visual Products
A professional framework for designing and deploying AR, VR, and 3D immersive shopping experiences tailored for high-value, visual-centric products.
Silverback DeFi Intelligence
Silverback DeFi provides 19 specialized x402-powered endpoints for real-time market data, technical analysis, and security audits on the Base chain.
Stock Study Equity Research
A high-fidelity equity research tool that transforms AI agents into senior investment analysts for deep ticker analysis and market consensus.
Happy Hues
A design-focused skill that provides curated, role-based color palettes to help AI agents generate beautiful and functional UI color schemes.
Continuity Framework
Continuity is a framework for AI agents to asynchronously reflect on sessions, extract structured memories, and surface relevant questions for future interactions.
Featured*
