Blog·Learning

Hermes Agent Memory with Supermemory

Configure the Hermes memory provider, inspect capture and recall, and separate work and personal context with explicit container settings.

By Dhravya Shah·4 min read

Persistent memory for Hermes agents

Hermes has built-in memory files and can use Supermemory as a native memory provider. The provider adds external recall, profiles and capture through Hermes's lifecycle. Check both capture and recall when setting up continuity for a project.

Start with the upstream provider documentation for your installed Hermes version. Provider behavior can change independently of the Supermemory API.

Install and configure

With Hermes installed, run:

pip install supermemory
hermes memory setup

Choose Supermemory and supply a key from the console. Keep the key out of source control. For manual configuration, the provider documents memory.provider and the SUPERMEMORY_API_KEY environment variable.

The coding plugin is free to use; hosted ingestion and retrieval consume usage under the current rate card. Free plugin access does not mean unlimited free API usage.

Check the capture path

The current upstream implementation captures completed user/assistant turns when enabled. It groups writes by session and a four-hour window and retries pending writes on later lifecycle events. Session-end handling retries pending work.

Retries are at-least-once: if a write succeeds but its response is lost, resubmission can append the turn again. A prolonged outage can also exceed the bounded retry buffer. Inspect results rather than assuming automatic capture means nothing can be lost or duplicated.

Use a synthetic conversation to verify what is stored. Include a rejected suggestion so you can check that a proposed plan is not treated as an accepted decision.

Configure scope explicitly

The documented default container is hermes. If profiles should be separated, set the supported identity template in $HERMES_HOME/supermemory.json:

{
  "container_tag": "hermes-{identity}"
}

Without that template or another explicit separation scheme, different Hermes profiles can use the same container. A profile name alone is not proof that external memory is isolated.

Optional multi-container mode allows named additional containers. Automatic capture and prefetch use the primary container; explicit tool calls can target an allowed additional container. Instructions guide the model's choice but do not guarantee perfect classification of work and personal content.

Verify the destination on an explicit save and the scope on a search. Repeat with an unrelated container as a negative test.

Use the exposed tools

The provider exposes save, search, forget and profile operations through supermemory-save, supermemory-search, supermemory-forget and supermemory-profile, with documented snake_case aliases. Check the returned record and scope before interpreting the final answer as successful recall.

Forgetting and source deletion are different operations. The memory lifecycle guide explains why hiding an extracted memory from normal retrieval is not a promise to erase every stored document or transcript.

Keep native files and external memory understandable

MEMORY.md and USER.md remain part of Hermes's native behavior. The current provider handles eligible additions through its memory-write hook. That does not establish a complete initial import or bidirectional synchronization of every edit and deletion.

If you change a local note, verify whether the external copy changes too. If you remove a fact, check both stores. A successful write hook does not establish that all existing file contents were imported when the provider was enabled.

Test before relying on a new session

Save a fictional project code name, end the session, and retrieve it from a new session under the intended scope. Then correct it and test the lifecycle. Check errors and pending writes if it is missing.

Compaction reduces active context. Captured source content may help recover details later, but capture and retrieval are separate steps. Include a compacted session in your recall test.

Self-host with an explicit data-flow boundary

The provider supports a base_url for a local or private Supermemory endpoint. Follow the self-hosting documentation and configure that endpoint before connection probes if those probes must remain local.

A local memory endpoint does not make the entire agent offline. Hermes's answering model, the memory service's configured model providers, connectors and other tools can still make external requests. Verify those dependencies for the deployment you intend to run.

Follow the Hermes integration guide and start with the cross-session test above. The useful result is a fact you can inspect, retrieve, correct and remove under the right scope.

Frequently asked questions

Does Hermes already have built-in memory?

Yes. Native memory files remain available alongside Supermemory's external provider. Check which source supplies context when testing recall.

Does every Hermes profile automatically get a separate Supermemory container?

No. The documented default is hermes. Use the supported identity template or another explicit configuration when separate containers are required.

Does a local memory endpoint make the whole agent offline?

No. Check the answering model, memory-service model configuration, connectors and tools as well as the memory endpoint.

  1. We're open sourcing the company brain. Here's how we designed the multiplayer harnessCompany Brain is now open source. A walkthrough of the multiplayer harness behind its Slack experience, from proactivity and memory boundaries to approvals and recovery.
  2. Jev changes a lot in memory & context engineering. Here's exactly how.We tested Jev across reranking, chunking, observation, and harness decisions. Here is where fast decision models help memory systems, where they cost more, and where they still fall short.
  3. I reverse-engineered Instinct's memory. Here's exactly how it worksInstinct keeps its memory as git-tracked markdown files, found with grep rather than vectors. Here is the whole system as far as black-box probing can reconstruct it, and how to rebuild it on supermemory in about 60 lines.
  4. An update to supermemoryWe've discontinued the supermemory company brain and Nova. Everyone who was charged has been refunded, our MCP and plugins continue to run, and we're going all in on the memory engine.
  5. Scaling Conversations: How Adapta Grew Usage Without Losing ContextAdapta added Supermemory as a persistent memory layer so every conversation keeps its context — letting the team scale usage without losing the thread.
  6. How Chatarmin Ditched RAG and Went Memory-Only with SupermemoryChatarmin replaced a heavy RAG pipeline with Supermemory's memory layer — cutting average AI response time from 40s to 12s and token usage by 40–50%.
  7. SMFS: making agentic retrieval 55% cheaper AND more accurateWe launched SMFS.ai (Supermemory Filesystem) a few weeks ago, with a simple bet: We can redesign the filesystem specifically for agents, with special files, structures, and commands that it can use for it's tasks. Today, SMFS is used by hundreds of companies to power their agents.
  8. Introducing Dynamic Dreaming: supermemory now connects the dots, for you.Dreaming is magical. TLDR: We're launching Dynamic Dreaming in supermemory today, which automatically works if you're using supermemory in any way - API, OpenClaw, Hermes agent, etc.
  9. Dear reader, we just made supermemory insanely cheap... the Context CloudWhen I first started building supermemory, I had one goal: To build the best memory system for AI. I would talk to customers, and find out that memory was not the only thing they needed - They were all setting up 7-8 different vendors at the same time.
  10. Introducing @supermemory/tools v2.0.0Today we're releasing v2.0.0. This release unifies the API across all agents sdk integrations from AI SDK to Mastra, makes conversation identity a first-class concept, and ships with memory saving on by default.
  11. Solving the Precision-Recall Tradeoff: Search Result AggregationWhen you're building memory for AI, search is your foundational layer. The way search generally works is straightforward: the user defines a query, and then sets a limit (top-K) on how many search results they want returned. Usually, this is set to 10 or 20.
  12. OpenClaw Memory Problems: Why It Forgets and How to Fix It (2026)TLDR: Today, we are releasing a new version of our openclaw plugin - https://github.com/supermemoryai/openclaw-supermemory. This post is going to be a bit technical, so bear with me (or bookmark for later!) In this post, I will talk about what we do about OpenClaw memory, and how we fix it.
  13. Stateful Coding Agents with Memory: Build Long-Running Agents (2026)We built a plugin for Claude Code and OpenCode that gives your coding agent persistent memory. It remembers your preferences, learns your codebase, and never loses context mid-conversation. The result is an agent you can run for months without starting over.
  14. Clawd / Molt bot's memory SUCKS. We gave it supermemory.I'm the founder of supermemory. Clawd/Molt bot is blowing up right now, with many, many use cases. I set it up, too, and have been using it through telegram. TLDR: just go to https://supermemory.ai/docs/integrations/clawdbot to set up supermemory for your clawd bot.
  15. Catch up with our UNFORGETTABLE Launch WeekOver the last year, one belief has guided almost everything we’ve built at Supermemory AI becomes meaningfully useful only when it remembers. Memory shouldn’t be something developers rebuild from scratch. It shouldn’t be fragile, expensive, or trapped inside a single tool.
  16. Empowering the Next Generation of Founders: Supermemory Startup ProgramIf there’s one thing we’ve learned while building Supermemory, it’s that most startups don’t fail because they didn't build features; they fail when infrastructure slows them down, or they built too slow.
  17. Building code-chunk: AST Aware Code ChunkingAt Supermemory, we're building context engineering infrastructure for AI. A huge part of that is dealing with code: ingesting repos, understanding structure, and making it searchable. The problem is that most code chunking solutions are terrible. We built code-chunk to fix this.
  18. Supermemory raises $3 million with the best memory engine for LLMsToday, I am excited to announce our first funding round to accelerate our mission of building an interoperable, scalable and reliable memory for LLMs and agents. Memory is one of the hardest challenges in AI right now.
  19. Mem0 vs Supermemory: Why Scira SwitchedScira AI moved its production memory layer from Mem0 to Supermemory. This is what failed, what improved, and how the team evaluated the two systems.
  20. Never Record Again: How Montra Uses Supermemory to Rethink Video CreationCampbell Baron, the founder of Montra, has been making videos since he was twelve. By thirteen, he was already doing brand work. Today, he’s betting on a very different future for creators: a world where recording is the exception, and most videos are generated from scratch.
  21. Unified Memory That Works Where You Work: Your Second Brain With SupermemoryHi everyone, I’m Dhravya, the founder of Supermemory. I want to start with a little story behind why this product means so much to me. You can also skip straight to what it is and how it works below.
  22. Supermemory just got faster on PlanetScaleWhat is Supermemory? Supermemory completes the missing part of the LLM puzzle: memory. Just as memory is crucial for human intelligence, it's essential for truly intelligent AI systems.
  23. Faster, smarter, reliable infinite chat: Supermemory IS context engineering.People are obsessed with prompts and prompt engineering. Sure, what you say is important, but what the model knows when you say it is the difference between a stateless text generator and an intelligent AI system. In short, context is the most crucial component.
  24. We solved AI API interoperabilityOne API to rule them all, One spec to find them, One library to bring them all and in the TypeScript, bind them. When we were building the Infinite Chat API, initially, we only supported the OpenAI format. This was fine, until a lot of our customers started asking for more.
  25. The Wow Factor of Memory - How Flow Used Supermemory To Build Smarter, Stickier ProductsOverview: Flow is a note-taking app built around a bold vision: to create a more personal, context-aware writing experience powered by AI. At the heart of this mission is memory.
  26. The UX and technicalities of awesome MCPsLast month, we launched the Supermemory MCP, mostly to test our own infrastructure and get some initial traction. It blew up. To my absolute surprise, the initial launch itself got half a million impressions (!!!). Then, we launched and got #2 on ProductHunt too.
  27. Architecting a memory engine inspired by the human brainLanguage is at the heart of intelligence, but what truly powers meaningful interaction is memory — the ability to accumulate, recall, and contextualize information over time. Large Language Models (LLMs) have mastered language, but memory remains their Achilles’ heel.
YOUR PRIVACY

Change or withdraw any time via Cookie settings in the footer. Read our cookie notice.