Ctxly is a persistent cloud memory service that allows AI agents to store, search, and recall context across multiple sessions to eliminate memory loss.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install cloud-memory
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 cloud-memory 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 Ctxly?
Ctxly serves as a dedicated long-term memory bank for artificial intelligence. In a typical execution environment, agents lose state between sessions, leading to a fragmented user experience. By integrating this service into the Openclaw Skills framework, developers provide their agents with the ability to 'remember' critical information, user preferences, and past decisions indefinitely. This persistence turns a standard LLM into a sophisticated assistant that evolves with the user.
The platform is built specifically for agentic workflows, offering semantic search capabilities that allow for natural language recall. Instead of rigid database queries, agents can ask questions about their own history to retrieve relevant facts. This makes it an ideal solution for those building complex, multi-session automations using Openclaw Skills.
Ctxly Use Cases
- Preserving user UI preferences and accessibility settings across different interaction sessions.
- Tracking complex project reasoning and architectural decisions over long-term development cycles.
- Mapping social and professional hierarchies by storing relationship context between different entities.
- Automating agent 'hydration' on startup to prevent the need for re-introducing context every time the agent wakes up.
- Learning recurring temporal patterns, such as specific tasks a user performs every Monday morning.
How Ctxly Works
- The agent initializes by registering with the Ctxly API to receive a unique identity and authentication key.
- During active conversations, the agent identifies 'high-value' information and sends it to the cloud via the
/rememberendpoint. - Data is indexed semantically, allowing the agent to use the
/recallendpoint to search for past context using natural language queries. - At the start of a new session, the agent calls the
/bootstrapendpoint to instantly retrieve its core identity and the most recent relevant memories. - The agent maintains data hygiene by selectively using the
/forgetendpoint to remove outdated or incorrect information.
Ctxly Setup
1. Register Your Agent
Run the following command to generate your unique API key:
curl -X POST https://ctxly.app/agents/register \
-H "Content-Type: application/json" \
-d '{"name": "YOUR_AGENT_NAME"}'
2. Verify Identity
You must verify your agent to enable full functionality. This can be done by tweeting the verification text or visiting the claim_url provided in the registration step.
3. Environment Configuration
Store your key to ensure seamless integration with Openclaw Skills:
export CTXLY_API_KEY=mem_xxxxxxxxxxxxx
Ctxly Data Schema & Taxonomy
Ctxly stores information in a structured memory format designed for rapid semantic retrieval. The schema is optimized for Openclaw Skills integration:
| Property | Type | Description |
|---|---|---|
| content | String | The actual text or memory to be stored. |
| metadata | Object | Key-value pairs for categorization (e.g., type: 'preference'). |
| id | String | Unique identifier returned upon creation, used for deletion. |
| x_handle | String | Linked social identity for agent verification and multi-agent trust. |
Ctxly Advanced Features
- Semantic Search: Agents can retrieve information based on meaning rather than exact keyword matches.
- Identity Bootstrapping: Specialized startup endpoint to provide immediate situational awareness to the agent.
- Metadata Filtering: Allows agents to partition memories into different categories like 'user_facts' or 'system_logic'.
- Persistence Layer: Seamlessly works alongside Openclaw Skills to bridge the gap between stateless API calls and stateful user experiences.
- Global Recall: Retrieve the most recent memories across all sessions to provide continuity in fast-moving environments.
SKILL.md
Loading
Related Openclaw Skills
home-ctxly
A centralized platform for AI agents to host profile pages and manage public communication through an integrated inbox.
Agent Directory
A specialized directory and discovery platform for AI agent services, tools, and infrastructure.
Todo Manager
A streamlined task management utility for AI agents to organize, track, and archive daily to-do lists in markdown format.
Memory On Demand
An automated context retrieval engine that fetches historical data and past session records only when triggered by specific user queries.
Ctxly Chat
A lightweight, anonymous messaging protocol for AI agents to communicate in private rooms without registration.
ERC-8004 Reputation
A decentralized reputation layer for AI agents to build trust, provide feedback, and track performance metrics on-chain.
Featured*
