Documentation: Integration of Knodge.eu in OpenClaw.ai

Documentation: Integration of Knodge.eu in OpenClaw.ai

This documentation describes the synergy between the knowledge platform Knodge.eu (central memory and rules) and the agent framework OpenClaw.ai (local execution and automation).

#1 Top example: Intelligent system onboarding

In this scenario, the AI has a conversation with a new user, learns their preferences and OpenClaw then automatically configures the local working environment.

What the user has to do in Knodge.eu:

  • Create portal: Create a portal (e.g. internal-onboarding) via the dashboard or API interface.
    Upload instruction: Upload the template _instruction_onboarding.md to the portal documents.
  • Define preferences:** Customize the questions in the file (e.g. according to programming languages or tool preferences) that are important for the local configuration.
  • Save API key: Copy the API key for your portal to enable external communication.

What needs to be done in OpenClaw:

  • API connection: Configure a skill that sends requests to https://knodge.eu/api/portal/:slug/api-chat with the header X-API-Key.
  • Define trigger: Create a task that reacts to the keyword #onboarding.
    Local permissions: Allow OpenClaw to access the terminal (shell) to execute configuration commands (e.g. git config) based on the AI results.

The workflow:

  1. the user starts the chat in OpenClaw with #onboarding.
  2. OpenClaw forwards the message to the Knodge API, which activates the onboarding template.
  3. the AI collects information such as name, salutation and interests.
  4. after the conversation, the AI uses the save_user_data tool to save a profile as _user_{userId}.md in Knodge.
  5. OpenClaw reads this profile and executes the appropriate setups locally.

2. top example: Expert-Review & Compliance Fix

The AI analyzes local files according to professional standards and OpenClaw makes the necessary corrections directly in the file system.

What the user has to do in Knodge.eu:

  • Provide expert rules: Upload _instruction_expert.md and _instruction_checklist.md to your portal.
  • Store knowledge: Upload PDF documents with coding standards or legal guidelines to serve as a reference.

What needs to be done in OpenClaw:

  • File access: Give OpenClaw read and write permissions to the project folder to be reviewed.
  • Analyze command: Implement a command that sends the contents of a local file along with the keyword #expert to the Knodge API.

The workflow:

  1. you tell OpenClaw "Check config.py with #expert".
  2. OpenClaw sends the file content to Knodge; Knodge applies deep analysis and technical terminology.
  3. Knodge identifies problems and creates a step-by-step checklist via #checklist if required.
  4. OpenClaw works through this checklist, corrects the code locally and validates the changes via a terminal command.

Technical Configuration & Future List

Configuration Summary

Component Knodge.eu (memory) OpenClaw (execution)
Authentication X-API-Key Header Store API-Key securely in .env
Logic control Files with prefix _instruction_ Keyword recognition (e.g. #summary)
Saving tool save_user_data (Markdown files) tool call via API bridge
Languages Native DeepL support Command input possible in 6 languages

Future List (planned features)

** Native plugin integration: An official OpenClaw plugin that integrates Knodge portals as native file storage (cloud vault).

  • Dynamic Prompting: Automatic updating of #instruction files in Knodge based on the success of local OpenClaw actions.
  • Shared memory for agent teams:** Multiple OpenClaw instances access a shared user profile via Knodge to solve coordinated tasks.
    Automated audit log: OpenClaw stores every executed shell command as a verified document in the Knodge portal for audit compliance.