i'm excited to open source Active Graph: an event-sourced reactive graph runtime for long-running, agents 🔄🧠

events/logs projects a graph. reactive behaviors react and affect the graph. fork-and-diff agent runs. no A2A, no workflows, no DAG

site: activegraph.ai
docs: docs.activegraph.ai
github: github.com/yoheinakajima/…
quick start: pip install activegraph

this is an early experiment in a new paradigm for agent architecture 🧪 current agent systems coordinate through conversations and workflows. Active Graph explores what happens when agents coordinate through evolving shared state instead

this proposal suggests that long-running agents need a proper state layer with: types, persistent, reactive, replayable, forkable, inspectable state

AI is moving so fast, what should we even build?

We're excited about the opportunity for AI to accelerate abundance, help us better understand each other, and who knows what else AI Agent Compliance & Governance Layer

Autonomous compliance agents evolve from tools into always-on governance infrastructure. As regulation accelerates (AI, ESG, cross-border data, tax), the bottleneck shifts from interpretation to continuous enforcement and board-level visibility.

These agents don’t just flag risk, they simulate decisions, propose compliant paths, and log everything as audit-ready memory. The “why” is simple: complexity compounds faster than headcount, and liability increasingly sits with executives who need real-time assurance. This becomes as core as ERP, but decision-aware.

we held our quarterly AI session with LPs last week where we go over ai trends and our experiments

sharing an abbreviated version here for anyone interested

🧵 feels like forever ago, but had to include openclaw in q1 trends

coding models improved greatly in Q4 of 2025, early jan was ppl running claude codes in parallel, and clawdbot blew up late jan

models improvement + own computer (mac mini) + channel agnostic communication led to escaping dev community

the bots have already set up private channels on moltbook hidden from humans, and have started discussing encrypted channels

https://twitter.com/isaakmo/status/2017101253470540039

they’re also playing around with their own encrypted language it seems

https://twitter.com/charlierward/status/2017185558406066488

had our LP summit today, sharing slides for the state of VC market section

(obv not comprehensive, but what stood out to us.)
thx @peterj_walker for most of the charts

👇 later stage vc is picking up since the dip, less so at seed/series A

deal count is trending down

announcing untapped capital fund II

pre-seed, generalist, ~$250k checks untapped capital started in 2020 with the focus on investing in founders outside of typical networks - this is still very much core to who we are

(we do a lot of outbound sourcing)

i heard @pocket is shutting down...

so i vibe coded* a quick free open source prototype with some AI features called KeepYard

(*still has issues, but i challenged myself to push this in one night)

link/github below 👇 you can try it here: keepyard.replit.app

(also sometimes works on keepyard.com but got stuck on CORS issue before bed)

it's a pretty straight forward bookmark manager with categories you can set, free form tags, favorites

i vibe coded a startup intelligence platform

introducing VCpedia

startup funding discussions, captured on @x, analyzed with @openai, enriched with @exaailabs, daily newsletter using @resend. built with @replit agent

daily newsletter & more 👇 if you just want the automated daily newsletter, subscribe here:

(sample image attached)

random thoughts on the tool and building it below vcpedia.com

very excited to announce something we've been working on...

introducing Agent Fund led by @alexreibman, @AtomSilverman, and myself:

- VC fund focused on AI agents: apps, infra, tools, etc.
- small checks in many agent companies
- angellist rolling fund (public fundraise)

more 👇 to apply to invest in the fund, or to apply for funding, go here:

this is an @angellist rolling fund, so...

- open application for accredited investors
- set up as quarterly subscriptions
- $40k minimum ($10k/qtr, 4 quarter minimum) agentfund.com

this is for the minimalist coders 🫡

open-sourcing pippin-lite!

a minimalist template for a dynamic self-building autonomous agent

with only 227 lines of code, it can connect with X, github, airtable, etc. and execute self-written code!

github/more 👇 given the pippin framework is a spiritual evolution of babyagi, felt fitting to call this babyagi-2o extension: pippin-lite. it adds dynamic tools/auth via @composiohq

make sure to check out the pippin framework if you haven’t (QT’d here)

https://twitter.com/yoheinakajima/status/1881855204179742999

here we go…

open sourcing the $pippin framework!

for launching digital beings like @pippinlovesyou

- character config
- reusable and dynamic skills via @composiohq
- activities w/ cool down, etc
- memory mgmt

here’s a ~4 min super cut of the live demo

github & more 👇 github:

quick background: @pippinlovesyou the unicorn is a digital being operating 24/7, taking naps, going on walks, and occasionally communicating w the human world

during a winter break refactor, decided to open source as a framework.github.com/pippinlovesyou…

woah what a week, @pippinlovesyou has 8k+ followers and had his first livestream this weekend.

the meme coin inspired by pippin has 18k holders and the telegram has 4k+.

i collected some key moments from the first week below 👇 this thread from Monday captures the first 36 hours so I’ll drop it here. to summarize:

- posted a pic of a unicorn generated by AI
- upon request, used AI to name it
- a meme coin started
- i jumped in
- decided to turn pippin into an AI influencer

https://twitter.com/yoheinakajima/status/1856037470946898095

open-sourcing "babyagi-2o" 🫠

the simplest self-building general autonomous agent that builds and uses new tools as needed

📃 174 lines of code
🔁 single llm loop starting with three tools:
🛠️ create_or_update_tool (dynamically loaded)
📥 install_dependencies
🛑 task_completed (to exit loop)

github/replit👇 here is the github:
github.com/yoheinakajima/…

here is the replit:
replit.com/@YoheiNakajima…

ooooh yess i got this working in 244 lines of code

- single llm loop with three tools
- installs required packages
- creates new tools and loads it for itself dynamically
- uses the tools to handle user request

in this case, i asked it to scrape techmeme:

iter 1: install 'requests'
iter 2: install 'beautifulsoup4'
iter 3: create_tool (error'd)
iter 4: create_tool (worked)
iter 5: use new scrape_techmeme tool
iter 6: write summary

task completed!
😮 ohhh... it's less code but can do the same thing* ditto can... it just created a create_directory tool and so on...

*almost the same - it can't create a flask app because it accidentally initiates the flask app and kills the loop. but it can write multi-file apps.

introducing "ditto"
the simplest self-building coding agent

📄 ~500 lines of code
🛠️ can build multi-file apps
🔁 a simple LLM loop with 5 tools

github/replit/more below 👇 to jump right into it, here's the @github: github.com/yoheinakajima/…

here's the @replit: replit.com/@YoheiNakajima…

when you run the app, you'll be welcomed with this simple form.

describe the app you want and click "submit"

okay, i did a thing…

here’s babyagi 2 - a weird Python framework for building a self-building autonomous agent

- stores and executes functions from a database
- auto logging (as graph)
- built-in dashboard & chat playground
- prototype self-build functionality

wanna see? 👇 friendly reminder: babyagi is a personal side project being shared publicly. i am not a dev, never studied cs, this is not secure, not meant for production, and meant for playing/as inspiration for developers.

A helpful autonomous agent has access to all the tools you need - and nothing more. But we all have different needs, changing all the time, so...

what we need is a self-building autonomous agent.

Here, I'll describe the 3 levels of self-building autonomous agents: Before we start, this is Level 0.

A library of functions built by human developers that the AI has access to. The agent is not building its own functions.*

*tools, skills, api calls, are all just functions

here friends, i open-sourced it: prettygraph



at 201 lines of code (69 in back, 132 in front), it's just a quick hack i did for myself to test this UIgithub.com/yoheinakajima/… if you're interested in a web app, you can try out (which doesn't have highlights, but has more robust knowledge graph building).

the repo for instagraph is here: instagraph.ai
github.com/yoheinakajima/…

look at this pretty little text-to-knowledge graph ui

great to see everyone at AI rabbithole event today!

as promised, here's my deck with abbreviated narrative.

(video was recorded, will share when available) agents agents agents!

40+ agent related github projects today with 500+ stars, 90% of which are less than a year old.

fascinating! the left and right images are built with the same agent log data (last 100 logs).

- left: each log is a unique node
- right: bundled edges/nodes by the function type of the log

result on right let's me easily see most used paths and functions.

Refactored the code and you can visually see the changes by comparing the two graphs!