Pinned Tweet
Introducing ThinkThen, a CLI and SDK for System One models like Jev.
ThinkThen adds fast, cheap, good decisions to agents and "ordinary" software.
Ten simple, easy-to-understand functions built in Rust (fast, memory-safe) and ported to 24 languages and databases.
ThinkThen adds fast, cheap, good decisions to agents and "ordinary" software.
Ten simple, easy-to-understand functions built in Rust (fast, memory-safe) and ported to 24 languages and databases.
2
0
6
868
A: Yes. The question and the curiosity behind it are the value.
That’s and the answer fed back into the model will enhance its ability to answer future questions.
That’s and the answer fed back into the model will enhance its ability to answer future questions.
I have had two glasses of wine so lets do a thought experiment:
It is 2030. A scientist gathers data on a black hole so far away that it will never impact earth and understanding it has no practical value. She then gives that data to an AI and asks the AI to explain the phenomena. The AI thinks for a while and returns an explanation that no human is able to understand.
Has knowledge been increased?
0
0
0
208
Can someone tell @thsottiaux my /slow idea? cc: @jxnlco
@thsottiauxo 5-hour limits and 3x usage during off-peak hours?
Or what about a /slow mode where the Codex harness slows you down during peak usage and ramps you back up as needed?
slow mode would tell you how much time you're being slowed down and how much usage bonus you're getting.
0
0
2
208
I'm excited about Jev for its speed, cost, ease-of-use, and it is better out of the box than the 0-shot models I have used in the past.
thinkthen.dev/blog/jev-is-tr…
thinkthen.dev/blog/jev-is-tr…
2
0
9
510
Introducing ThinkThen. Code asks a model one bounded question about text and gets a typed answer back: yes, no, one option, or a number. Then it acts.
Ten functions. One Rust engine. 24 bindings. MIT.
Supports System-One models like Jev.
Ten functions. One Rust engine. 24 bindings. MIT.
Supports System-One models like Jev.
ThinkThen lets code act on what text means. Simple 10 function SDK that runs on System One models like Jev. Built in Rust with 24 bindings, a command-line tool, and SQL support in DuckDB, SQLite and PostgreSQL. MIT licensed. x.com/i/article/2105…
2
0
6
7,769
ThinkThen lets code act on what text means. Simple 10 function SDK that runs on System One models like Jev. Built in Rust with 24 bindings, a command-line tool, and SQL support in DuckDB, SQLite and PostgreSQL. MIT licensed. x.com/i/article/2105…
3
0
3
8,187
“Flash-level” local frontier at ~27 t/s (~500 prefill) thanks to @antirez on my m5 128gb.
However, integrated harness/inference/sampling loop might be the true unlock.
It lightens the load on the model and corrects its work much more efficiently.
Check out the [upto] injected control op in the replies. I can imagine many more features to lighten the burden on the model.
Eliminating http overhead and inspectable/steerable tool calls (basically intellisense for LLMs) might make these models smarter and faster at doing real work than much bigger/better models.
However, integrated harness/inference/sampling loop might be the true unlock.
It lightens the load on the model and corrects its work much more efficiently.
Check out the [upto] injected control op in the replies. I can imagine many more features to lighten the burden on the model.
Eliminating http overhead and inspectable/steerable tool calls (basically intellisense for LLMs) might make these models smarter and faster at doing real work than much bigger/better models.
I'll tell you one thing that shows a fundamental advantage of the "agent handles the LLM inference" in the specific case of local single user agents/LLMs. You can stop the tool and report an error as soon as in the first parameters of the tool calling you detect a problem. No need to wait for a long "content" generation that will fail immediately after.
0
0
4
359
`pi reload` doesn't reload auth, but it does reload extensions.
And you can make an extension that reloads auth.
And you really don't need to know how any of it works.
You just ask pi, and it figures it out.
And you can make an extension that reloads auth.
And you really don't need to know how any of it works.
You just ask pi, and it figures it out.
1
0
2
348
Biomedical agents can understand the treatment landscape for KRAS G12C-positive NSCLC by chaining together 7 BIOMCP commands.
The below terminal-based screencast demonstrates the BioMCP commands that:
* Orients across 1,741 known variants and 84 clinical trials
* Pulls CIViC evidence for G12C
* Maps 5 drugs targeting KRAS (sotorasib, adagrasib, and 3 investigational)
* Reviews sotorasib's FDA approval history and safety profile
* Checks 2,465 real-world adverse event reports from FDA FAERS
* Finds 62 recruiting trials for KRAS G12C NSCLC
* Retrieves the latest research on resistance mechanisms
7 entity types. 10+ data sources. Under two minutes.
Drugs, evidence, safety, trials, resistance -- the complete picture from a single CLI.
Full walkthrough with every command and output:
biomcp.org/blog/kras-g12c…
Documentation:
biomcp.org
GitHub Code repo:
github.com/genomoncology/…
#BioMCP #PrecisionOncology #KRASG12C #NSCLC #AIAgents #ClinicalGenomics
The below terminal-based screencast demonstrates the BioMCP commands that:
* Orients across 1,741 known variants and 84 clinical trials
* Pulls CIViC evidence for G12C
* Maps 5 drugs targeting KRAS (sotorasib, adagrasib, and 3 investigational)
* Reviews sotorasib's FDA approval history and safety profile
* Checks 2,465 real-world adverse event reports from FDA FAERS
* Finds 62 recruiting trials for KRAS G12C NSCLC
* Retrieves the latest research on resistance mechanisms
7 entity types. 10+ data sources. Under two minutes.
Drugs, evidence, safety, trials, resistance -- the complete picture from a single CLI.
Full walkthrough with every command and output:
biomcp.org/blog/kras-g12c…
Documentation:
biomcp.org
GitHub Code repo:
github.com/genomoncology/…
#BioMCP #PrecisionOncology #KRASG12C #NSCLC #AIAgents #ClinicalGenomics
0
1
5
721
Pi really is the best coding agent when you want to have sovereignty over your workflow.
I was frustrated that pi wasn’t handling my VTT workflow because it would treat the newlines by @WisprFlow as separate commands.
Just described the frustration and it fixed it with a ’t know if it’s the best solution (playing with minimax 2.7 which came up with it) but easily tweakable.
Works great and no PR required that @badlogicgames needs to worry about.
I was frustrated that pi wasn’t handling my VTT workflow because it would treat the newlines by @WisprFlow as separate commands.
Just described the frustration and it fixed it with a ’t know if it’s the best solution (playing with minimax 2.7 which came up with it) but easily tweakable.
Works great and no PR required that @badlogicgames needs to worry about.
2
4
79
8,607
BioMCP has been expanded to 12 entities and 29 trusted sources. ChatBots and Coding Agents can use BioMCP tool either as a command line interface (CLI) or model context protocol (MCP) server.
4
4
25
41,632
Ian Maurer 🧬 retweeted
Introducing kuva: A scientific plotting library in rust, along with cli binary with the option to plot directly into the terminal.
Feel free to drop me some feedback as an issue on the repo
github.com/Psy-Fer/kuva
crates.io/crates/kuva/0.…
Feel free to drop me some feedback as an issue on the repo
github.com/Psy-Fer/kuva
crates.io/crates/kuva/0.…
Yea your plotting library is cool, but can it do this?
(yes, that's plotting in the terminal, with just ascii, utf-8, and ansi)
11
33
407
26,539
I wish /status told me:
% tokens remaining for week
% hours remaining for week
Doing this math all the time makes my head hurt.
🙏 @trq212 @thsottiaux
% tokens remaining for week
% hours remaining for week
Doing this math all the time makes my head hurt.
🙏 @trq212 @thsottiaux
0
0
3
265