The Filesystem for AI Agents
San Francisco
Joined November 2025
- Tweets32
- Following8
- Followers444
- Likes56
Mesa retweeted
I'm so excited to officially share what we've been cooking up at Mesa: the most powerful filesystem ever built for AI agents.
The dirty secret of every "production" AI agent today: the filesystem is held together with duct tape.
Teams are stitching together S3, GitHub, sandbox-local disks, and homegrown diff logic to give their agents something resembling persistent, versioned storage. None of it works.
S3 isn't designed for parallel agents - concurrent agent writes silently overwrite each other.
GitHub has the semantics but rate-limits you into the ground at agent scale and doesn't give you filesystem ergonomics.
Sandbox disks vanish the moment the container dies.
And your agents don't want to git clone and git push anyway. ๐ง๐ต๐ฒ๐ ๐๐ฎ๐ป๐ ๐๐ผ ๐ฟ๐ฒ๐ฎ๐ฑ ๐ฎ๐ป๐ฑ ๐๐ฟ๐ถ๐๐ฒ ๐ณ๐ถ๐น๐ฒ๐. ๐๐ถ๐ธ๐ฒ ๐ฒ๐๐ฒ๐ฟ๐ ๐ฝ๐ฟ๐ผ๐ด๐ฟ๐ฎ๐บ ๐ฒ๐๐ฒ๐ฟ ๐๐ฟ๐ถ๐๐๐ฒ๐ป.
So we built the missing layer.
Mesa is a durable, POSIX-compatible filesystem with version control built in. Branches, diffs, history, rollback, access control โ every primitive a codebase has, for any file type, at agent scale. You mount it. Your agent uses it like a normal filesystem. We handle the rest.
Private beta is live. Link in comments.
Mesa retweeted
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents.
Every team building agents eventually hits the same wall: where do the files live?
Not the chat history, the actual artifacts the agent works on.
> The contracts your agent redlined
> The claim files it updated
> The 200-page audit report it edited overnight while you were asleep
Today those documents live in a sandbox that dies in 30 minutes, an S3 bucket where concurrent writes clobber each other, or a GitHub repo that was never built to absorb agent-scale traffic.
So we built Mesa.
The world's first POSIX-compatible filesystem with built-in version control, designed from the ground up for agents. You mount it into your sandbox like any other filesystem. Your agent reads and writes files normally. Behind the scenes every change is versioned, branchable, reviewable, and rollback-able โ like a codebase, for any file type.
Mesa provides
โ Branches so agents work in parallel without locking
โ Durable storage that survives sandbox death
โ Sparse materialization so massive document sets load instantly
โ Fine-grained access control per agent
โ Full history for human review and audit
Design partners are running Mesa in production across legal, healthcare, GTM, business ops, and coding agents.
Private beta is open: link in the comments
Mesa retweeted
You can review 100K lines of code for free with @mesa_dot_dev. They also hit #1 in recall on bug-fix PRs โ so the free tier isn't a compromise.
Mesa retweeted
Really enjoyed this conversation with @og_doctourist.
As engineers, @olvrgln and I always obsessed over the tech.
As founders, we had to learn entirely new skills:
- choosing the right bets early
- earning trust and community
- moving fast without breaking people
@mesa_dot_dev is trying to replace Github. That is absolutely nuts!!
Ex-Microsoft engineers move to San Francisco, link up at @spc, and decide to reinvent the entire software development lifecycle for AI agents.
They build a code review product that actually understands your full codebase, and land their first 20-50 customers just by calling their friends. No cold outreach.
The company today:
- Seed stage startup in San Francisco
- AI-native source forge for agents & dev teams
- 20-50 early customers onboarded through network alone
- Hiring aggressively for top engineers
The best shows are art (storytelling) and science (frameworks). This has both off the charts. Enjoy EP5 of No Fixed Route - the first ever tech podcast filmed in a @zoox.
My 5 takeaways with @bwarrn in the replies
Mesa retweeted
This month, Mesa increased our internal code output by 2.4x.
Today weโre open-sourcing the tool that made it possible: Agent Blame.
Everyone has opinions about which AI agents/models are best for codegen. Almost nobody has hard data on their own codebase.
So we built a way to attribute every AI-written line of code to the agent + model that generated it โ using Git, with zero workflow changes.
Now we can answer questions like:
โข merged lines per dollar
โข bugs per LOC
โข which agents actually work best
Bonus: seeing AI attribution in PRs let us review code ~50% faster.
We open-sourced it so every team can make data-driven AI decisions.
And yes โ our top agent was Claude Code ๐. Congrats to @bcherny and the Anthropic team.
