pumaDB

Nice launch. The memory I’d trust most is not just facts, but attempts: what the agent tried, why it failed, and which tool or write it was allowed to use next.If a memory is wrong or stale, does pumaDB show who or what wrote it and let builders expire or correct it?

I have seen teams, including my own, avoid memory because setting up a database or vector store feels too heavy for early workflows. How do you decide what should be saved as memory and what should stay out? Can developers inspect and clean up memory when an agent saves something wrong or outdated?

I really like the pitch for this. I too have run into this problem, and not everyone has the time, energy, and willpower to research all of the different skills and scaffolding and harness options to make your own ideal memory layer. Speeding up that whole process, to me, is a very empowering thing to give to people.

The exact same idea I was thinking about.
is it possible to integrate it in chatgpt or claude web interface?

The transcripts example resonates. I've got the same problem with a different domain. Iterative LLM workflows where each session starts cold means re-feeding context that should persist. For "what would agents remember automatically": the thing I'd actually pay for is automatic capture of failed attempts and why they failed. Most memory tools focus on remembering successful artifacts. The harder and more valuable thing is remembering the dead ends so the agent doesn't try the same broken approach next session.MCP-first feels right for the dev audience. Curious if you're planning to expose the memory as a queryable index later or keeping it strictly key/notes.

I built pumaDB because I kept running into the same problem with AI agents: they do useful work, then the useful context disappears into chat history, local files, Notion, GitHub, or some custom setup.

I wanted something simpler.

pumaDB gives agents a shared memory they can read and write through MCP or a server-side API. You can use it for things like project context, research notes, transcripts, reusable snippets, preferences, decisions, task state, and things already tried.

It is intentionally lightweight. It is not trying to replace Postgres, vector search, or your production database. It is for the smaller but very common problem of giving agents a reliable place to remember useful context across sessions and tools.

A simple example: I moved transcripts from my last 23 videos into pumaDB. Now I can ask Claude, ChatGPT, Codex, or Conductor to summarize, repurpose, or search that same content without copying it between tools.

Would love feedback from anyone building with agents:
- What do you currently use for agent memory?
- Do you prefer MCP, API, or both?
- What would you want agents to remember automatically?
- What would make you trust a shared memory layer?