The AgentNotes skill captures and logs OpenClaw task execution states, generating clean hourly or daily SparkNotes summaries to keep developers and stakeholders informed.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agentnotes
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 agentnotes 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 AgentNotes for OpenClaw?
The AgentNotes skill is a powerful logging utility designed to integrate your AI agents directly with the AgentNotes platform. Instead of storing massive amounts of raw log lines indefinitely, this skill allows agents to send high-level, human-readable job summaries that are automatically rolled up into concise SparkNotes. This provides stakeholders and developers with an elegant, executive-level overview of agent performance, task status, and operational health.
By incorporating this integration into your workspace, you leverage the broader ecosystem of Openclaw Skills to maintain perfect visibility into automated workflows. Whether your agents are replying to communication channels, running routine cron jobs, or executing complex multi-step processes, the AgentNotes skill ensures every run is cleanly documented, easily searchable, and summarized without log bloat.
AgentNotes for OpenClaw Use Cases
- Post-Execution Summaries: Automatically log a high-level summary after an agent completes a task, such as responding to channel messages or processing email sweeps.
- Cron Job Monitoring: Keep track of scheduled background tasks and heartbeats, ensuring scheduled workflows are running as expected.
- Error & Failure Tracking: Capture and isolate run failures with detailed error outputs to debug operational issues instantly.
- Multi-Step Workflows: Log complex sequences step-by-step to visualize long-running agent reasoning and actions.
How AgentNotes for OpenClaw Works
- Initialization: The agent starts a run or initiates a task logging context using the local scripts provided by the skill.
- Execution & Event Logging: As the agent executes its functions, it logs specific milestone steps, messages, or intermediate results.
- Rollup Submission: Once the task finishes, the final summary payload is dispatched to the configured AgentNotes endpoint.
- SparkNotes Aggregation: The AgentNotes platform aggregates these logs hourly or daily, converting verbose technical execution details into clean SparkNotes.
AgentNotes for OpenClaw Setup
To get started with this skill from the Openclaw Skills registry, follow these steps:
- Configure AgentNotes: Head to your AgentNotes dashboard, create a new agent, copy its unique slug, and retrieve your API key from the Connect tab.
- Install the Skill: Run the installation script from your repository:
./install-skill.sh
- Update Configuration: Add the environment variables to your
~/.openclaw/openclaw.jsonfile:
{
"skills": {
"entries": {
"agentnotes": {
"enabled": true,
"env": {
"AGENTNOTES_API_KEY": "an_your_api_key_here",
"AGENTNOTES_AGENT_ID": "your-agent-slug",
"AGENTNOTES_BASE_URL": "https://your-app.vercel.app"
}
}
}
}
}
- Verify the Installation: Restart your OpenClaw session and run the verification script:
node {baseDir}/scripts/verify.mjs
AgentNotes for OpenClaw Data Schema & Taxonomy
The AgentNotes skill organizes logging payloads through structured JSON schemas before sending them to the API. Below is the metadata and schema setup:
| Parameter | Type | Required | Description |
|---|---|---|---|
AGENTNOTES_API_KEY |
String | Yes | Unique authentication token (an_...) issued by AgentNotes. |
AGENTNOTES_AGENT_ID |
String | Yes | The unique slug matching the agent created on the dashboard. |
AGENTNOTES_BASE_URL |
String | No | Custom self-hosted application URL (defaults to production endpoint). |
Task Execution Payloads
Logs sent from scripts dynamically build the following taxonomy:
summary: A plain-English, executive summary of the action taken.message: Detailed execution steps or intermediate output.step: Category label representing the active workflow segment (e.g.,inbox,tools).failed/error: Boolean flags and error descriptors captured during execution failures.
AgentNotes for OpenClaw Advanced Features
- Manual Multi-Step Session Control: Programmatically orchestrate step-by-step runs by exporting a run ID (
AGENTNOTES_RUN_ID) and completing the run at a later execution context. - Custom Endpoints: Supports self-hosted or private cloud deployments of AgentNotes via the optional
AGENTNOTES_BASE_URLconfiguration. - Resilient Error Logging: Allows explicit logging of failures with custom error descriptors to isolate SMTP, network, or third-party API issues immediately.
- Seamless Openclaw Skills Integration: Integrates natively with other modules to ensure task tracking is unified across telegram bots, cron triggers, and custom toolchains.
SKILL.md
Loading
Related Openclaw Skills
Mneme
Mneme is a lightweight, plain-text persistent memory protocol that enables AI agents to maintain cross-session decisions, facts, goals, and preferences safely.
Excalidraw Diagram Generator
An automated visualization engine that designs, structures, and exports clean, modern system architecture flowcharts and diagrams.
Global Travel & Work Planner
A one-stop bilingual travel and work guide generator supporting 48 countries that outputs dual-language Word (.docx) documents with fresh, web-fetched information.
AIPPT
The AIPPT skill enables AI coding agents to programmatically generate, customize, and compile professional slide decks using topics, documents, or custom Markdown outlines via CLI automation.
88Cha Bidding Search
An AI-ready command-line interface tool for searching and tracking real-time tendering and bidding announcements from 88Cha.
Venus Tracker
An automated project tracking execution skill designed to synchronize Feishu multi-dimensional tables (Bitable) and streamline daily reporting for ERP developments.
Featured*
