Alim

18. ceo @agentmashdev. lead swe @ astana hub.

ex-director of an IT-solutions business, projected ARR $150K in 4 months, $10k in the first 2 weeks.

[Damir I @iskakovdev]

humans resolved 96.1% of the conflicting merges in 12,299 ai-written PRs

the agent opens the PR. you clean up after it

i wrote a script that finds your own conflicted merges and who fixed them, plus 3 ways to make it cheaper. https://t.co/CgNqr0XW58

[Alim @alimwtf]

PR pairs from two different coding agents on one repo: 41.7% hit a merge conflict. Same agent twice: 19.8% (Xu et al., arXiv, July 2026).

How to run Claude Code and Codex on one repo, with tested commands.

How many agents do you run on one repo? https://t.co/pZUQVRst11

[Alim @alimwtf]

PR pairs from two different coding agents on one repo: 41.7% hit a merge conflict. Same agent twice: 19.8% (Xu et al., arXiv, July 2026).

How to run Claude Code and Codex on one repo, with tested commands.

How many agents do you run on one repo? https://t.co/pZUQVRst11

[Alim @alimwtf]

run claude code and codex on one repo long enough and they'll edit the same file.

here's a repro, the manual fix, and how agentmash warns your agent before the edit. https://t.co/1SWiAzanRx

[Alim @alimwtf]

🤣🤣

[dvb @notdvb]

are you fucking kidding me?

https://t.co/La3RH0f4f0

[Alim @alimwtf]

@iskakovdev lfg 🔥

[Alim @alimwtf]

yessir

[Damir I @iskakovdev]

let your team code simultaneously 🔥

multiplayer for ai agents is in early access now

Claude Code, Codex, Cursor, Gemini and any open source agent in one repo

[Damir I @iskakovdev]

let your team code simultaneously 🔥

multiplayer for ai agents is in early access now

Claude Code, Codex, Cursor, Gemini and any open source agent in one repo

[Alim @alimwtf]

every developer on your team runs an AI agent. none of them can see the others.

agentmash fixes that:
- your agent is warned before it edits a file a teammate's agent is changing
- see what every agent is doing, live
- ask a teammate's agent. nothing runs until they approve

[Alim @alimwtf]

the era of opus generated motion design has begun

[nicco @0xnicc0]

Opus 5.5 just made After Effects optional, and nobody in motion design is ready for it

I had to watch this three times…

Drop the PDF below into Opus 5.5, you’ll thank me later.

Then save the article on the Opus 5.5 best practices and usage for motion design.

[Alim @alimwtf]

we're in closed beta. request early access at https://t.co/P6CDzsOj3Z

[Alim @alimwtf]

every developer on your team runs an AI agent. none of them can see the others.

agentmash fixes that:
- your agent is warned before it edits a file a teammate's agent is changing
- see what every agent is doing, live
- ask a teammate's agent. nothing runs until they approve

[Alim @alimwtf]

that's why we're building @agentmashdev
its a tool that allows agents to see the context of what other developers are doing in real-time. Apply for early access: https://t.co/P6CDzsOj3Z

[Alim @alimwtf]

in one study of agent-written pull requests, 79.4% were open at the same time as another one in the same repo.

four in ten repos had any overlap, and nearly all of it came from the same agent tool.

[Alim @alimwtf]

thats why I'm building https://t.co/P6CDzsOj3Z, check it out

[Alim @alimwtf]

developers avoid stepping on each other by talking. "i'm in the billing code today."

agents can't. they have no channel to each other. so the more agents your team runs, the more often two of them edit the same thing blind.

[Damir I @iskakovdev]

If you run more than one coding agent, each one keeps its own rules, skills and MCP config.

Here's how to put Claude Code, Codex, Cursor and Gemini on one setup, then connect all of them to agentmash.

Every block is copy-paste. Bookmark it https://t.co/dw5zx8exO9

[Damir I @iskakovdev]

🆓 🆓 GLM 5.3 Flash is available in agentmash, free through OpenCode

Run it next to Claude Code, Codex and Gemini with the same skills and rules, and hand a chat from one to another

Let your team try it on agentmash 🔥

[Alim @alimwtf]

@thedelost if you run agents in parallel with your dev team, use https://t.co/P6CDzsNLer to give real-time context of what your agents are doing to your teammates' agents

[Alim @alimwtf]

@benyuls and if your developer team uses agents to work in parallel, use https://t.co/P6CDzsNLer to give them a real-time context layer