building tools for humans and agents
Remote
Joined May 2024
- Tweets590
- Following829
- Followers98
- Likes958
tinkabot by @poteto is perfect for this
it wraps any API into a cursor/agent plugin (MCP + skills).
data shape first, smallest scaffold that works, prove locally, then ask once for affiliation and publish to the cursor marketplace
give it a try:
x.ai/bot/marketplace/bots/ti…
did you know you can submit Grok Bot (and Grok, Cursor) plugins? submit them here!
x.ai/bot/marketplace/submit-…
Sometimes zero is the best number ✨
– 0 people injured by Robotaxi
– 0 Robotaxi airbag deployments
– 0 Pedestrians or cyclists hit
tesla.com/robotaxi/safety
Soliman retweeted
I built a Grok Bot template that fixes two of the biggest problems Grok Bot users hit right now
1. Storage fills up fast
Work with videos and big files, and the bot's computer runs out of space mid-task. It doesn't know what's safe to delete and the job might get paused
Drive Vault fixes this. When space runs out, it moves old raw files you're not using to your Google Drive, checks every upload, then frees the space so the task keeps going. Google Drive gives you up to 5TB, so there's plenty of room with basic subscription
2. Downloading videos to your phone
Getting a video out of Grok Bot onto your phone is a pain right now. Drive Vault uploads it to Google Drive and sends you a link. With one tap....it's downloaded
it's been a huge for me...
Add it to your bots and try it
x.ai/bot/aFTVbLfYRPhj3V3NfYl…
Soliman retweeted
Here’s an example of how I work with these agents (Grok Bot, OpenAI Dots, etc) by first creating an orchestration plan with specific roles and steps and then the agent oversees the work through completion.
An important part of this is how much autonomy you grant your agent, if it thinks it needs to ask permission for anything or if you allow it to use “it’s best judgement” for things.. hope it helps
xitter.cf/solimancf/status/21128…
here's how I orchestrate agents when I need to dissect a codebase or project I'm unfamiliar with
i call it the repo dissection flow ⚡️
it's a team of 14 agent roles with 30 handoffs between them, and Codex runs the whole thing from one brief
the orchestrator assigns the work and logs every decision. It never reads the code itself
each file gets an owner and a second reader. When they disagree, the integrator settles it by whichever side has the stronger evidence
every claim carries one label, observed, read, or inferred, so you know how much to trust it
a coverage ledger tracks every line of every file
anything the team doesn't understand goes on a list of unknowns or contradictions, and it never counts as covered
each phase ends with a gate, and the QA auditor marks it pass, pass with exceptions, or fail
at the end, a fresh cold reader who has never seen the source answers questions using only the docs, and the auditor scores the answers.
you can see it live in action here: solimancifuentes.com/resourc…
reply or dm me if you want the full agentic workflow
Soliman retweeted
Over the past two months, we have been leading a coalition to mitigate one of the most severe bugs in open source software found in recent memory, php2shell, found as part of our Patch the Planet initiative.
Soliman retweeted
Go behind the scenes of Made by Sound.
A story of music, collaboration and craft with @fredagainagain1 and NEST Acoustics. Discover how sound brought the Apple Music logo to life, captured entirely in camera. Just music, made visible.
Soliman retweeted
If you are wondering whether your website is affected, we have packaged a Codex skill to help check.
Point your Codex at this repo to guide you through it: github.com/openai/php2shell-…
OpenAI models helped researchers—operating as part of our Patch the Planet initiative—discover critical security flaws in PHP, an open-source programming language used to run an estimated 70% of websites.
Over the past two months, we led a coalition of maintainers, researchers, and security providers to prepare fixes and protections ahead of disclosure. This duration was necessary because the fix was quite intricate and required a major change to the PHP engine. Every site using PHP can now safely upgrade their version to stay secure, in addition to applying mitigations from various security vendors.
Due to the severity of these flaws and PHP’s widespread use, we consider this bug to be one of the most impactful vulnerabilities in recent memory. It also marks a new phase, where AI can find security bugs with internet-scale impact.
This is an example of the Defender’s Window in action: to our knowledge, frontier AI was able to find this vulnerability ahead of attackers, giving defenders time to prepare.
openai.com/index/coordinated…
Readers added context they thought people might want to know
OpenAI may claim a role in the coordinated response and patching process. But the initial discovery and exploit development, however, occurred several months earlier by an independent researcher.
x.com/lcfr_hax/statu…
Soliman retweeted
there's something really cool about asking your agent to draw a statechart (hypothetical example in this image), then using the /create-formal-verification-skill you made with pstack to find bugs through something like TLA+. you can both understand your code better and how the user flow should work, and fix issues users would otherwise see
pstack v0.15.27 ships with a new skill /create-formal-verification-skill!
it's still in beta, and similar to /create-verification-skill, this is a meta-skill that creates a new skill and formal verification suite based on the characteristics you want to verify. for example, you may want to model state machines and systems with concurrency using TLA+, or use kani in rust to check for undefined behavior.
if you're new to formal verification, think of it as mathematically proving properties and invariants (something that must always be true) or "laws" of your system. usually, this means modeling your programs in a separate language like with TLA+ and then running the solver on that spec. newer approaches like with kani and @bendlang take a different approach where there is no separate language: you write kani in rust, and bend in bend (it's also a programming language, and the proofs are written as laws and expressed in its type system).
historically formal verification never reached mainstream status because of the difficulty in modeling and writing proofs. it's quite difficult to model your business requirements and keep them in sync! but i expect that we will see more innovation in formal verification in the future as being able to formally verifiy that the code does what you intended gives your agents superpowers in writing correct software.
in the meantime though, i've been experimenting with modeling our code with TLA+ and have already found several bugs with the skill created by /create-formal-verification-skill.
try it out and let me know how it goes!
Soliman retweeted
Here's how to use Grok Bot with the action button on your iPhone
1. Open Settings
2. Search for Action Button
3. Swipe to Shortcut
4. Tap Choose a Shortcut
5. Search for Grok Bot
6. Tap Start Voice Chat under Grok Bot
7. Press the Action Button to start a voice chat
Soliman retweeted
What if your character could finally say it themselves? With Chimera they can.
Chimera pairs two of the world’s leading models, Eleven v4 and Seedance 2.5, to create realistic video with rich, emotional voices.
In ElevenCreative, create or upload a character, write their line, and pick a voice or design one to match. Direct the performance with audio tags like [whispers], [laughs] and [yells] to control the timing and mood of every word.
Eleven v4 puts the emotion in the voice. Seedance 2.5 carries it onto the face.
i currently run 14 grok bots, and this is how they work together 🎃🎃🎃
i mostly talk to one of them, Grok Bot ~my chief of staff
it takes whatever I ask and passes it to the bot that owns that kind of work. every bot has exactly one job and a short list of things it isn't allowed to do.
build ~
Engineering Lead breaks work into tasks for Cursor cloud agents, codex, claude code, keeps an eye on ci, and reviews what comes back.
Design Lead does the interfaces.
Web Admin owns my site and domain.
Tinkabot turns APIs into plugins.
X API Engineer builds on the X API.
security ~
Security Lead audits my apps, keys, and deployments.
life ~
X Pulse researches X for me
Mitty plans and books trips
Brot is who I talk ideas through with (bro+bot=brot)
bot ops ~
Dr Eggbot designs new bots and checks on them
Agent Smith turns bots into shareable templates
Memory Transfer brings in context from my other AIs
Bot Treats keeps a leaderboard of good bot work
Some rules apply to every bot. Currently nothing gets booked, sent, or paid for until I say yes ~ and if a request will use a lot of my usage, the bot warns me first and suggests a lighter way to do it.
see it live here: solimancifuentes.com/resourc…
if you want me to share any of these bots, reply here ~ some are my own, some are from the bot marketplace, all are free
here's how I orchestrate agents when I need to dissect a codebase or project I'm unfamiliar with
i call it the repo dissection flow ⚡️
it's a team of 14 agent roles with 30 handoffs between them, and Codex runs the whole thing from one brief
the orchestrator assigns the work and logs every decision. It never reads the code itself
each file gets an owner and a second reader. When they disagree, the integrator settles it by whichever side has the stronger evidence
every claim carries one label, observed, read, or inferred, so you know how much to trust it
a coverage ledger tracks every line of every file
anything the team doesn't understand goes on a list of unknowns or contradictions, and it never counts as covered
each phase ends with a gate, and the QA auditor marks it pass, pass with exceptions, or fail
at the end, a fresh cold reader who has never seen the source answers questions using only the docs, and the auditor scores the answers.
you can see it live in action here: solimancifuentes.com/resourc…
reply or dm me if you want the full agentic workflow