# How to Install and Use Supermemory in OpenClaw - Tool: OpenClaw - Difficulty: Beginner - Time required: 10 minutes - Last updated: March 2026 ## TL;DR To use Supermemory in OpenClaw, run `clawhub install supermemory`, get your Supermemory API key from supermemory.ai, configure SUPERMEMORY_API_KEY in OpenClaw's config, and start saving and retrieving information across sessions from OpenClaw chat. Supermemory gives OpenClaw a persistent knowledge layer — facts, notes, and context you save in one session are searchable and retrievable in any future session. ## Why Supermemory Is the Persistent Knowledge Layer for OpenClaw Every OpenClaw conversation starts fresh — the AI agent has no memory of what you discussed last week, what decisions you reached last month, or what research you gathered last year. This context amnesia is one of the most limiting aspects of AI agent workflows. Supermemory solves it by providing a structured, searchable knowledge store that persists independently of any single conversation. What you save to Supermemory in one session is available in every future session, transformed into context that OpenClaw can reason over. The use cases are broad and compound in value over time. You might save meeting notes to Supermemory after each project call, so that OpenClaw can always answer questions about past decisions by searching your saved memory. You might save research findings as you conduct them, building a personal knowledge base that grows with every session. You might save your preferences, code style guidelines, or recurring context about your team and projects — so that OpenClaw always has that context available without you re-explaining it from scratch at the start of every conversation. Supermemory is particularly valuable in combination with other OpenClaw skills that generate useful outputs. ByteRover might produce a data analysis summary, Sag might generate a validation report, Larry might collect a weekly review — Supermemory preserves all of these outputs as searchable knowledge rather than letting them disappear when the conversation session ends. Over weeks and months, your Supermemory knowledge base becomes a meaningful record of work done, decisions made, and insights gathered — all queryable through OpenClaw chat. ## Before you start - OpenClaw installed and running (see openclaw.ai for installation instructions) - ClawHub CLI available in your terminal (bundled with OpenClaw) - A Supermemory account at supermemory.ai to generate your API key - Terminal/command line access to run the clawhub install command ## Step-by-step guide ### 1. Install the Supermemory Skill via ClawHub Open your terminal and run the clawhub install command to add Supermemory to your OpenClaw instance. ClawHub downloads the supermemory skill package, verifies its integrity against the registry, and registers it with your local OpenClaw configuration. Installation takes 10-25 seconds. When it completes, the terminal shows a confirmation message with the installed version and the required configuration key: SUPERMEMORY_API_KEY. The skill is installed but inactive until you configure the API key — any attempts to save or retrieve memory before configuration will return a clear error pointing to the missing credential. After installation, run `clawhub info supermemory` to see the full list of available configuration options, including optional settings for default tags, memory categories, and retrieval preferences that you can configure later to customize how Supermemory organizes and surfaces your saved knowledge. ``` clawhub install supermemory ``` **Expected result:** Terminal shows 'supermemory@X.X.X installed successfully. Required config: SUPERMEMORY_API_KEY'. The skill appears in `clawhub list` output. ### 2. Get Your Supermemory API Key Navigate to supermemory.ai and sign in or create an account. Supermemory offers a free tier that provides enough storage for personal use and initial exploration — check their pricing page for current tier details. Once logged in, navigate to your account settings or developer dashboard and look for 'API Keys' or 'Access Tokens'. Click to create a new API key and give it a descriptive name like 'OpenClaw Integration'. Supermemory generates the key immediately — copy it to your clipboard now. Your Supermemory API key is the authentication credential for all save and retrieve operations from OpenClaw. It is scoped to your Supermemory account, so all memories saved through OpenClaw are associated with your Supermemory account and accessible through the Supermemory web interface as well as through OpenClaw. Keep the key private — it grants full read and write access to your Supermemory knowledge store. If it is ever compromised, you can revoke it from the Supermemory dashboard and generate a new one without losing your saved memories. **Expected result:** You have a Supermemory API key copied to your clipboard, ready to add to OpenClaw's configuration. ### 3. Configure the Supermemory API Key in OpenClaw Add your Supermemory API key to OpenClaw's configuration using the `clawhub config set` command or by adding it to your `~/.openclaw/.env` file. Both methods write to the same secure configuration store and the key will be available to the supermemory skill on next reload. After setting the key, run `clawhub reload` to apply the configuration without a full OpenClaw restart. Verify the key is correctly saved by running `clawhub config get SUPERMEMORY_API_KEY` — you should see a masked version of your key confirming it is present. Optionally, configure the supermemory skill's YAML file at `~/.openclaw/skills/supermemory.yaml` to set default behavior for saves and retrievals — such as a default space or default tags to apply to all saved memories from OpenClaw. These defaults make saving consistent without requiring you to specify organizational details in every prompt. ``` # Option A: Add to ~/.openclaw/.env SUPERMEMORY_API_KEY=your_supermemory_api_key_here # Option B: Set via clawhub config command clawhub config set SUPERMEMORY_API_KEY your_supermemory_api_key_here # Verify the key is saved clawhub config get SUPERMEMORY_API_KEY # Apply changes clawhub reload ``` **Expected result:** `clawhub config get SUPERMEMORY_API_KEY` returns a masked value. Running `clawhub status` shows supermemory as 'active'. ### 4. Test Supermemory by Saving and Retrieving Information Open OpenClaw chat and test the Supermemory skill by saving a piece of information and then retrieving it in a new prompt. This two-step test confirms that both the save operation (write to Supermemory API) and the retrieve operation (read from Supermemory API) are working correctly. For the save test, provide a clear title, some content to save, and optionally some tags to categorize the memory. The supermemory skill formats this into a Supermemory API call and returns a confirmation with the memory ID once saved. For the retrieve test, reference the title or tags you used in the save and ask OpenClaw to search Supermemory for it. A successful retrieval returns the saved content with its title, tags, and the timestamp of when it was saved. Once the basic save/retrieve cycle works, you can explore more sophisticated retrieval patterns — semantic search across multiple saved memories, filtering by date range or tags, and combining retrieved context with fresh information from other skills. **Expected result:** OpenClaw confirms the memory was saved to Supermemory with a memory ID. A follow-up retrieval prompt returns the saved content with its title and tags intact. ### 5. Build a Knowledge Management Workflow with Supermemory With Supermemory working, establish a knowledge management habit that compounds in value over time. The most effective use of Supermemory is consistent, structured saving — tagging memories with topics, projects, or dates so that retrieval is precise rather than requiring broad semantic searches across everything you have ever saved. A practical workflow is to end each OpenClaw session with a save step: summarize the key insights, decisions, or outputs from the session and save them to Supermemory with the project name and date as tags. Over weeks, this builds a rich retrievable record of your work that OpenClaw can query at the start of future sessions. Supermemory also excels as the output target for other OpenClaw skills. After ByteRover produces a data analysis, save the summary to Supermemory. After Sag runs a validation pattern, save the findings. After Larry completes a weekly review, save the key points. Each skill output that would otherwise disappear at session end becomes a permanent, searchable record. RapidDev helps teams set up Supermemory as a shared organizational knowledge base within OpenClaw, particularly for teams who want consistent memory organization standards across multiple users and projects. ``` # Configure Supermemory defaults in ~/.openclaw/skills/supermemory.yaml supermemory: default_space: "openclaw" # default Supermemory space for saves default_tags: ["openclaw"] # tags automatically added to every save max_results: 10 # max memories returned per search (default: 5) retrieval_mode: "semantic" # semantic | exact | hybrid (default: semantic) include_metadata: true # include timestamps and tags in retrieval auto_save_summaries: false # auto-save session summaries (optional feature) ``` **Expected result:** Supermemory returns a structured summary of all saved memories tagged 'project-phoenix', organized by date, giving you immediate context for continuing work on the project. ## Best practices - Save information to Supermemory with clear, searchable titles — 'Project Name — Decision Type — Month Year' format is highly retrievable compared to vague titles like 'notes from today'. - Use consistent tag taxonomy across all your saves — decide on a standard set of project names, topic categories, and date formats before you start saving, and apply them consistently. - Store your SUPERMEMORY_API_KEY only in your OpenClaw `.env` file or via `clawhub config set` — never in chat messages, shared files, or version-controlled repositories. - Build the habit of saving important session outputs before ending an OpenClaw conversation — information not saved to Supermemory is lost when the session context is cleared. - Combine Supermemory with ByteRover, Sag, and Larry: save the outputs from those skills to Supermemory after each run, building a persistent record of analyses, validations, and reviews. - Set `retrieval_mode: hybrid` in your skill config for the best balance of exact and semantic search — this reduces missed retrievals when you cannot remember the exact wording of a saved memory. - Periodically review and clean up your Supermemory knowledge base through the web interface at supermemory.ai — over time, outdated or redundant memories can dilute search result quality. - For team use, establish shared tagging conventions before multiple people start saving to a shared Supermemory space — inconsistent tags make shared knowledge harder to search reliably. ## Use cases ### Persistent Research Knowledge Base Save research findings, key facts, and source citations to Supermemory as you gather them across multiple OpenClaw sessions. Build a knowledge base that grows over time and is searchable by topic, date, or tag — so that information gathered weeks ago is instantly retrievable in today's conversation. Prompt example: ``` Save this to Supermemory under the title 'AI regulation EU — March 2026 summary' with tags 'AI regulation', 'EU', 'compliance': The EU AI Act enforcement deadline for high-risk AI systems was confirmed for Q2 2026, requiring conformity assessments for systems in healthcare, critical infrastructure, and biometric identification. Source: European Commission official announcement, March 2026. ``` ### Project Context and Decision Log Log key project decisions, architectural choices, and team agreements to Supermemory as they happen. When you return to a project after time away, OpenClaw can retrieve the decision history from Supermemory and immediately provide context without requiring you to re-read documentation manually. Prompt example: ``` Save to Supermemory under 'Project Phoenix — Architecture Decisions — March 2026': We decided to use PostgreSQL over MongoDB for the primary datastore because the team has stronger SQL expertise and the relational data model fits our customer-account relationship structure. Reviewed and approved by the engineering team on March 15. ``` ### Personal Preferences and Standing Instructions Store your preferences, style guidelines, and standing instructions in Supermemory so that OpenClaw can retrieve them at the start of conversations and apply them consistently — without you needing to repeat context in every session. Prompt example: ``` Search Supermemory for my stored writing style preferences and code formatting guidelines, then apply them to help me write the technical spec for the new authentication module. ``` ## Troubleshooting ### OpenClaw returns 'supermemory: authentication failed' or '401 Unauthorized' when saving or retrieving Cause: SUPERMEMORY_API_KEY is not set in OpenClaw's configuration, is set under an incorrect variable name, or the key value was copied with extra whitespace. Solution: Run `clawhub config get SUPERMEMORY_API_KEY` to check if the key is present and non-empty. If it returns empty, re-run `clawhub config set SUPERMEMORY_API_KEY your-key-here`. Open `~/.openclaw/.env` to verify there are no spaces around the equals sign or extra characters in the key value. Then run `clawhub reload`. ``` # Check and re-set the API key clawhub config get SUPERMEMORY_API_KEY clawhub config set SUPERMEMORY_API_KEY your_supermemory_api_key_here clawhub reload ``` ### `clawhub install supermemory` fails with '429 Too Many Requests' or a timeout error Cause: ClawHub's skill registry enforces rate limits during peak usage. This is a registry infrastructure issue unrelated to Supermemory. Solution: Wait 2-5 minutes and retry. If a partial installation is blocking a clean retry, run `clawhub install supermemory --force` to clear the incomplete state. ``` # Force reinstall clawhub install supermemory --force ``` ### Searching Supermemory returns empty results even though I know I saved relevant information Cause: The search query does not semantically or textually match how the memories were saved — different wording, different tags, or different spelling than what was stored. Solution: Try different search phrasing, broader keywords, or the exact title you used when saving. Check the Supermemory web interface at supermemory.ai to see how your memories are labeled and what titles and tags were applied — this reveals the exact terms you need to search for from OpenClaw. ### Supermemory saves are successful but memories are not appearing in the Supermemory web dashboard Cause: Memories may be saved to a different Supermemory space than the default view in the web dashboard, or there is a short propagation delay after saving. Solution: Check the Supermemory web dashboard for all available spaces, not just the default view. If the memory was saved to a custom space configured in your supermemory.yaml `default_space` setting, select that space in the web UI. Newly saved memories may take up to 30 seconds to appear in the web dashboard. ## Frequently asked questions ### How do I install Supermemory in OpenClaw? Run `clawhub install supermemory` in your terminal. Then get a free API key from supermemory.ai, configure it with `clawhub config set SUPERMEMORY_API_KEY your-key-here`, and run `clawhub reload`. You can immediately start saving and retrieving information from OpenClaw chat once the skill is active. ### What does Supermemory do in OpenClaw? Supermemory gives OpenClaw a persistent long-term memory. Information you save in one OpenClaw session — notes, research findings, project decisions, preferences — is stored in your Supermemory knowledge base and retrievable in any future session. This solves the context amnesia problem where every OpenClaw conversation starts with no memory of previous interactions. ### How do I configure the Supermemory API key in OpenClaw? Run `clawhub config set SUPERMEMORY_API_KEY your-key-here` from the terminal, or add `SUPERMEMORY_API_KEY=your-key-here` to your `~/.openclaw/.env` file. Run `clawhub reload` after either method. Verify it is saved with `clawhub config get SUPERMEMORY_API_KEY`. ### Does Supermemory in OpenClaw share information across sessions? Yes — that is Supermemory's primary purpose. Information saved via the supermemory skill in one OpenClaw session is stored in your Supermemory cloud account and retrievable in all future sessions. The persistence is not limited to OpenClaw — you can also view and manage your saved memories in the Supermemory web interface at supermemory.ai. ### Why does searching Supermemory return empty results even though I saved the information? The search query wording does not match the saved memory's title, content, or tags closely enough. Try rephrasing with different keywords, use the exact title you saved the memory with, or check the Supermemory web dashboard to see the exact labels and tags applied. The `hybrid` retrieval mode in skill config helps by combining exact and semantic matching. ### Can RapidDev help set up Supermemory as a team knowledge base with OpenClaw? Yes — RapidDev can help configure Supermemory for team use within OpenClaw, including establishing shared tagging conventions, integrating Supermemory with other OpenClaw skills for automatic knowledge capture, and setting up consistent memory organization structures across team members. Reach out to RapidDev for team knowledge base configuration support. ### Is there a limit to how much I can save in Supermemory through OpenClaw? Supermemory's storage limits depend on your plan. The free tier provides a limited number of memory items and storage capacity — check supermemory.ai for current free tier limits. If you plan to use Supermemory extensively through OpenClaw for professional research or team knowledge management, consider a paid Supermemory plan to avoid hitting free tier limits. --- Source: https://www.rapidevelopers.com/openclaw-integrations/supermemory © RapidDev — https://www.rapidevelopers.com/openclaw-integrations/supermemory