Konteks for Openclaw

A persistent context layer that connects your AI agent to Konteks for shared memory, task management, and project organization.

jamesalmeida
v1.0.2
Feb 3, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install konteks

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install konteks 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 Konteks?

Konteks provides a robust context layer for Openclaw Skills, enabling agents to maintain long-term memory across sessions. It acts as a bridge between the agent's internal logic and a user's Konteks account, allowing for the synchronization of tasks, notes, and organizational structures. By implementing Konteks, developers can move beyond transient chat sessions to create agents that understand user preferences, remember past decisions, and actively manage a user's digital workspace.

This skill is essential for anyone building sophisticated agents that require a reliable source of truth and a way to interact with real-world productivity tools. By integrating this into Openclaw Skills, the agent gains the ability to see what the human sees, ensuring that context is never lost even if the local environment is reset or the agent is accessed from a different device.

Konteks Use Cases

  • Storing agent memories and user preferences across multiple devices to ensure consistent behavior.
  • Managing tasks and to-do lists directly within the Openclaw Skills framework.
  • Organizing agent output into structured projects and areas for better information architecture.
  • Tracking long-term decisions and learnings that persist beyond a single conversation.
  • Syncing agent-generated notes with the Konteks web and iOS apps for human-in-the-loop workflows.

How Konteks Works

  1. The agent initializes the session by pulling the most recent memories and context from the Konteks API to restore its state.
  2. During a conversation, the agent utilizes a hybrid approach, keeping fast scratchpad notes in local files while sending significant decisions to the Konteks database.
  3. Proactive heartbeats trigger the agent to perform inbox triage, automatically filing new items into the correct projects when the destination is obvious.
  4. The agent monitors for due or overdue items, alerting the human user to urgent priorities during its routine checks.
  5. Every significant learning, preference, or project note discovered is upserted to the cloud, ensuring it remains available for future Openclaw Skills sessions.

Konteks Setup

To integrate Konteks with your agent, follow these steps:

  1. Sign up for an account at konteks.app.
  2. Navigate to Settings and generate a new API Key.
  3. Add the required environment variable or update your configuration for Openclaw Skills:
skills:
  konteks:
    apiKey: "sk_your_key_here"
    agentId: "my-agent" # optional
  1. Ensure the KONTEKS_API_KEY is available in your agent's environment.

Konteks Data Schema & Taxonomy

Konteks organizes data into four primary entities to maintain a clean context layer within Openclaw Skills:

Entity Description Key Fields
Agent Context Persistent memories, decisions, and preferences category (decision, preference, learning), key, value
Items Tasks, notes, or hybrid entries visible in the app title, body, priority, due_date, tags, smart_list
Folders Structural organization for Projects and Areas name, folder_type, icon, goal
Daily Plans Daily schedule and task summaries task_ids, summary, calendar_events

Items created by the agent are automatically tagged with a source of ai for easy filtering.

Konteks Advanced Features

  • Proactive inbox triage during heartbeats to keep the user's workspace organized without manual intervention.
  • Hybrid memory architecture that balances high-speed local storage with cloud-based persistence for Openclaw Skills.
  • Multi-agent support via unique agent_id identifiers, allowing different agents to maintain separate or shared memory pools.
  • Smart list management to categorize items into Inbox, Anytime, or Someday based on urgency and context.
  • Decision logging that captures the rationale behind agent actions, making the AI's behavior more transparent and auditable.

SKILL.md


Loading

Related Openclaw Skills