SessionMind
SessionMind — Claude That Knows You. Every Session.
94 followers
SessionMind — Claude That Knows You. Every Session.
94 followers
Set your context once — who you are, what you do, what you're working on — and Claude remembers you across every session. No more re-explaining yourself. Unlike Claude's built-in memory, SessionMind is fully transparent. Your context is visible, editable, and under your control. You decide what Claude knows. Features: persistent context, full conversation history, model selection (Haiku, Sonnet, Opus), streaming responses, markdown rendering. Free during beta. No credit card required.
Hey PH! I'm Srini, and I built SessionMind because I was frustrated with the same thing every Claude power user deals with: starting every conversation from zero.
I use Claude for everything, product strategy, code reviews, consulting prep, writing. But every new session, I had to re-explain who I am, what I'm building, and what I need. It felt like meeting the same person for the first time, every single day.
SessionMind fixes that. You write your context once, and Claude remembers you across every session. Full conversation history, model selection (Haiku, Sonnet, Opus), and your context is always visible and editable. No black box.
Free during beta, no credit card required. I'd love your feedback on what to add next. Thanks for checking it out!
Congrats on the launch, Srinivas. Japan-based founder here, using Claude/Codex heavily.
One Japan-specific thought: for Claude memory tools here, the blocker may not be “can Claude remember me?” but whether memory stays safe and understandable across messy real work contexts: Japanese docs, mixed JP/EN prompts, client-specific notes, internal specs, and project boundaries.
Japanese solo founders and small teams often work across Notion, GitHub, Slack/LINE, and local docs, so the local trust angle I’d test first is not just “Claude that knows you,” but “Claude memory with explicit boundaries, so client/project context does not quietly blend together.”
That could matter a lot for consultants, dev shops, and solo builders here who want continuity, but are cautious about persistent AI memory.
@wakuta Sorry for the late reply. your comment was buried under. I responded to every comment but yours. Thank you, this is exactly the kind of signal I was hoping to get on launch day.
The point about context blending is sharp. "Claude that knows you" is table stakes. "Claude memory that respects project and client boundaries explicitly" is the harder and more valuable problem, especially for consultants and dev shops handling multiple clients.
Japan-specific context, mixed JP/EN prompts, and the Notion/GitHub/Slack/LINE stack you described — that's a real workflow I had not fully modeled. The local trust angle you're raising is something I want to build toward deliberately, not bolt on later.
Would love to stay in touch as this evolves. Are you building in this space too
@srinivas_narra Thanks, Srinivas. I’m not building a Claude memory product directly right now. I’m a Japan-based solo iOS developer using Claude/Codex heavily across product work, QA, App Store/TestFlight workflows, and messy JP/EN project context.
So my angle is more operator-side: what makes persistent AI memory feel safe and useful in real client/project workflows. I agree the boundary/trust layer is the product here, not just recall.
Happy to stay in touch as you evolve it.
@wakuta That operator-side perspective is actually more valuable to me right now than another builder's take. You're describing exactly the workflow SessionMind needs to hold up in — multi-client, mixed language, real project pressure.
iOS dev across App Store, TestFlight, and messy JP/EN context is a great stress test. If you're open to it, I'd love to loop you in as I build out the boundary and trust layer. Early feedback from someone living this daily would be worth a lot.
Let's stay in touch.
@srinivas_narra @Srinivas_narra Absolutely, happy to be looped in.
The main value I can add is as a real workflow stress-test user, not as another memory-feature theorist.
My daily context is messy: iOS/App Store/TestFlight work, Claude/Codex, Notion/GitHub/Slack/LINE, JP/EN prompts, and different client/project boundaries. The failure mode I care about most is not “memory forgot something,” but “memory brought the wrong context into the wrong project.”
If useful, I’d be happy to share concrete boundary cases and react to early builds as you shape that layer.
I really like that the context is always visible and editable. A lot of these wrappers try to hide the 'magic,' but as a dev, I need to see exactly what’s being sent to the model to avoid hallucination. Clean approach @srinivas_narra
Very nice! Few questions:
lots of OSS projects in this space. what do you guys think about that?
does this run with claude.ai or claude code (or co-work)?
do you guys think claude won't roll this feature out themselves?
@abhishekmathur On OSS competition:
Great question. There are OSS tools like TypingMind and LibreChat that offer similar wrapper functionality. The difference is SessionMind is opinionated about simplicity. No setup, no self-hosting, no config files. Sign up, set your context, chat. For power users who want to tinker, OSS is great. For people who just want Claude to remember them without a weekend DevOps project, that's us.
On Claude.ai / Claude Code / Cowork:
SessionMind runs as a standalone web app powered by the Claude API. It's not a plugin for claude.ai or an extension for Claude Code. Think of it as your own Claude workspace where you control the context layer. You pick the model (Haiku, Sonnet, Opus), set your context, and everything persists across sessions.
On Claude rolling this out themselves:
They might! I have seen them do it before. Claude already has built-in memory. But it's a black box. You can't see exactly what it remembers, you can't edit it precisely, and you can't port it. SessionMind gives you full control. Your context is a text field you own. If Claude eventually builds something better, great. Until then, we're solving the problem today.
I hope that answers some/if not all of your questions. Feel free to give it a spin. No credit card req. And you piggyback on my Anthropic Token. Thanks for engaging.
-Srini