solo setups | ai researcher building @kollectivexyz 444
everywhere
- Tweets 8,182
- Following 170
- Followers 20,143
- Likes 22,052
Pinned Tweet
REA is dangerous with Opus 5.5
It lets your agent inspect apps you have no source for
But finding the code is not the same as proving the behavior
I broke down how to get from a black box to a build you can actually test ↓
36
7
94
30,269
rari retweeted
Holy f*ck, Opus 5.5 replaced After Effects
It's f*cking insane…
Drop the PDF below into Opus 5.5 and see what it makes
Then read the full guide on motion design in the article below ↓
Opus 5.5 can make a video
The hard part is making the next one better
That takes a studio, not a prompt
I broke down how to build it ↓
67
276
3,740
262,415
Holy f*ck, Opus 5.5 replaced After Effects
It's f*cking insane…
Drop the PDF below into Opus 5.5 and see what it makes
Then read the full guide on motion design in the article below ↓
Opus 5.5 can make a video
The hard part is making the next one better
That takes a studio, not a prompt
I broke down how to build it ↓
67
276
3,740
262,415
rari retweeted
I turned my Opus 5.5 into a motion design studio
This is f*cking insane...
I asked them to arrange a test drive of the Tesla Roadster
Drop the PDF below into Opus 5.5 and see what it makes
Then read the motion engineering guide below ↓
Opus 5.5 can make a video
The hard part is making the next one better
That takes a studio, not a prompt
I broke down how to build it ↓
19
4
69
5,788
I turned my Opus 5.5 into a motion design studio
This is f*cking insane...
I asked them to arrange a test drive of the Tesla Roadster
Drop the PDF below into Opus 5.5 and see what it makes
Then read the motion engineering guide below ↓
Opus 5.5 can make a video
The hard part is making the next one better
That takes a studio, not a prompt
I broke down how to build it ↓
19
4
69
5,788
rari retweeted
REA + Opus 5.5 is one of the wildest engineering stacks right now...
And most people will use it to generate a confident guess
That paid feature isn't magic
If you are allowed to inspect it, you can trace its behavior and build your own version
one feature → REA evidence → behavior spec → clean-room rebuild → real tests
Send this single prompt to Opus 5.5 + REA
Then save the article below on becoming a REA engineer
REA is dangerous with Opus 5.5
It lets your agent inspect apps you have no source for
But finding the code is not the same as proving the behavior
I broke down how to get from a black box to a build you can actually test ↓
6
1
51
1,956
rari retweeted
REA is dangerous with Opus 5.5
It lets your agent inspect apps you have no source for
But finding the code is not the same as proving the behavior
I broke down how to get from a black box to a build you can actually test ↓
36
7
94
30,269
rari retweeted
Opus 5.5 + an AI memory stack is an insane reverse-engineering setup
Most agents can describe an app. The hard part is proving how one feature behaves, rebuilding it, and not losing the evidence when the chat resets
I wrote a 13-page field guide for that exact workflow:
Authorized target → Controlled probes → Evidence ledger → Behavior contract → Tested rebuild → Fresh-session verification
Send this to your Opus 5.5 ↓
Your agent remembers the deadline
It may also remember the one that got replaced
More history won't tell it which one still counts
I broke down the AI memory stack: what to save, what to load, what to recheck, and when to forget ↓
13
9
92
3,336
instinct users spend over $1,300 a month through the agent according to their founder
is your site ready for agents trying to buy?
We just solved attribution for AI Agents.
Introducing Journeys.
AI agents are visiting your product. Now you can see what they’re trying to do, what converts, and what’s making them churn.
With Journeys, you can:
- See the moment AI agents visit your product
- Follow the paths they take
- Track purchases, installs, and completed setup
- Pinpoint where agents get stuck and whats blocking them
- Review fixes prepared for you and publish them with a click
- Watch how those changes improve conversion
Turn agent traffic into revenue.
Try Journeys at tryscope.com
Paid partnership (ad)
11
1
78
2,520
Agent builders know this loop too well:
Call → Parse → Branch → Call again
What caught my eye with TypeLLM:
The next decision can depend on the last one inside a single request
Incident triage → Assess severity → Consider rollback if critical → Draft a status update
One API call. Typed results. Less glue code
Chat models return messages.
Decision models return decisions.
Workflow models execute workflows.
TypeLLM introduces the first workflow model. Run an entire workflow in a single API call.
Simpler, faster, cheaper.
Try it now: typellm.ai/examples/ticket-w…
Paid partnership (ad)
13
4
82
2,723
EntelligenceAI just open-sourced AgentPlay, a local replay tool for coding-agent sessions
Files read →Commands retried → Errors hit → Context lost
This is f*cking useful...
Run it on one old session
Self-improving harness starts by seeing what actually happened
Introducing AgentPlay 🚀
The future of coding is self improving coding agent harnesses that help agents avoid making the same mistakes over and over.
Turn your existing Claude Code, Codex, Cursor and Copilot sessions into interactive maps and replayable timelines.
See what your agent explored, where it failed, what it retried and how it reached the final change
Free, local and open source.
GitHub: github.com/Entelligence-AI/a…
Paid partnership (ad)
18
4
88
2,606
Opus 5.5 + an AI memory stack is an insane reverse-engineering setup
Most agents can describe an app. The hard part is proving how one feature behaves, rebuilding it, and not losing the evidence when the chat resets
I wrote a 13-page field guide for that exact workflow:
Authorized target → Controlled probes → Evidence ledger → Behavior contract → Tested rebuild → Fresh-session verification
Send this to your Opus 5.5 ↓
Your agent remembers the deadline
It may also remember the one that got replaced
More history won't tell it which one still counts
I broke down the AI memory stack: what to save, what to load, what to recheck, and when to forget ↓
13
9
92
3,336
rari retweeted
These prompts for Opus 5.5 and Fable 5.1 are insane for motion engineering
I picked 8 GitHub repos that can work as one stack:
Remotion, GSAP, Three.js, React Three Fiber, Lottie, SVGO, FFmpeg and Puppeteer
The trick: Remotion owns the clock. The other tools only join when a shot needs them
Brief → Storyboard → Frame-accurate motion → Render → Visual QA
No competing timelines. No installing 8 packages just because they're on the list. No "done" without checking the export
Pick Opus or Fable, replace the bracketed fields and send it the prompt ↓
Opus 5.5 can make a video
The hard part is making the next one better
That takes a studio, not a prompt
I broke down how to build it ↓
14
6
67
2,903
Most AI agents still work in one person's chat
Kylon puts them in the team's workspace
Claude Code, Codex and other agents can share context, hand off work and connect to 3,000+ services, from Gmail and Slack to GitHub
This is what an actual agent team looks like
Introducing Kylon: The first company harness for your humans, agents and tools.
Whether you’re using Codex, Claude, Devin, Hermes, Gemini CLI, OpenClaw, Muse, Instinct, Grok Bot, Dots, Perplexity Computer, Cursor, Antigravity, Microsoft Autopilot, Amp, Droid, Kiro, Cline, OpenCode, Replit Agent, Lovable, v0, Gmail, Slack, Apollo, Supabase, Apify, Treg, Neon, Vercel, Databricks, Novita, RadixArk, GitHub, Seedance, Fish Audio, Plaud, OpenRouter, Gemini, Google Sheets, TikTok Shop, Outlook, Google Drive, Instagram, Airtable, Google Calendar, Asana, HubSpot, Corgi, SharePoint, Google Tasks, Notion, Google Workspace, Telegram, LINE, Zoho Mail, Meta Ads, Google Search Console, Fathom, ClickUp, Reddit, Granola, Google Docs, OneDrive, tl;dv, Amplitude, Dropbox, YouTube, Ahrefs, Xero, HighLevel, Google Analytics, Google Slides, Monday, QuickBooks, ElevenLabs, Linear, PostHog, LinkedIn, Runway, Zoho Bigin, Stripe, Exa, Cal_com, Box, Google Meet, TikTok, Wise, Mercury, Deel, Replit, Resend, Zoom, Fireflies, Typeform, Canva, Facebook, Google Admin, Hostinger, Google Maps, Retell AI, UptimeRobot, Google Photos, Excel, Cloudflare, Beehiiv, Vapi, 1Password, RunPod, Unipile, Read AI, Figma, Microsoft Teams, Trello, Salesforce, Jira, Confluence, Miro, Coda, Todoist, Basecamp, Wrike, Zendesk, Intercom, Freshdesk, ServiceNow, Gorgias, Mailchimp, ActiveCampaign, MailerLite, SendGrid, Postmark, Customer_io, Gong, Clay, Greenhouse, FreshBooks, Zoho Books, Brex, Ramp, PandaDoc, Google Ads, Microsoft Ads, Pinterest Ads, Semrush, Moz, Mixpanel, Google Tag Manager, Microsoft Clarity, Tableau, Metabase, Snowflake, Google BigQuery, ClickHouse, Wix, Contentful, Sanity, Bubble, Make, GitLab, Bitbucket, Sentry, PagerDuty, Grafana, New Relic, Docker Hub, DigitalOcean, Render, Firebase, Prisma, Algolia, Cloudinary, LaunchDarkly, Postman, Midjourney, Veo, GPT-Image2, NanoBanana, Flux, Kling, MiniMax, Replicate, HuggingFace, DeepSeek, Mistral AI, Ollama, Firecrawl, Eventbrite, Webex, Toggl, Harvest, Tailscale, Mural, Discord, TickTick, Calendly, Attio, Higgsfield, Square, Wispr Flow, Readwise, Bright Data, Google Places, FastMoss, PageSpeed Insights, Kalodata, Cerebras, SerpApi, DataForSEO…
Kylon enables them to work better, together.
Paid partnership (ad)
20
2
66
80,277