Yann Leflour
@yleflour.me
Engineer and maker at 💜
- Exploring software craftsmanship with AI @ https://pairprog.io
- Writing about it @ https://blog.pairprog.io
- Exploring software craftsmanship with AI @ https://pairprog.io
- Writing about it @ https://blog.pairprog.io
Pinned
Yann Leflour
@yleflour.me
· Feb 18
Fabrice Bernhard & Yann Leflour - Swift to React Native: AI-Driven Transpilation | RNLConf 2024
YouTube video by Theodo UK
www.youtube.com
Finally! Our React Native London talk with @fab-ber.bsky.social on AI assisted migrations is available online
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
3. I've invested in learning how to use this tool in my field of expertise. Holy cow, it's transformational but will be a plague for human intent, knowledge, and creativity.
There are basically only two positions in the debate about AI.
1. I’ve barely invested any time in learning how to use it effectively. AI sucks.
2. I’ve invested in learning how to use this tool. Holy cow, it’s transformational.
Position #1 has lower barriers to entry.
1. I’ve barely invested any time in learning how to use it effectively. AI sucks.
2. I’ve invested in learning how to use this tool. Holy cow, it’s transformational.
Position #1 has lower barriers to entry.
October 17, 2025 at 12:16 PM
3. I've invested in learning how to use this tool in my field of expertise. Holy cow, it's transformational but will be a plague for human intent, knowledge, and creativity.
It's July 2025 and prompting is still a skill
A skill I have a hard time teaching
Some people just get it, but most struggle
And no, LLMs are not very good at prompting themselves without another prompt to guide them
A skill I have a hard time teaching
Some people just get it, but most struggle
And no, LLMs are not very good at prompting themselves without another prompt to guide them
July 29, 2025 at 12:13 PM
It's July 2025 and prompting is still a skill
A skill I have a hard time teaching
Some people just get it, but most struggle
And no, LLMs are not very good at prompting themselves without another prompt to guide them
A skill I have a hard time teaching
Some people just get it, but most struggle
And no, LLMs are not very good at prompting themselves without another prompt to guide them
If you are tired as me of working with native path and fs libraries in NodeJS, you might like my
www.npmjs.com/package/@sy... NPM package.
A library to finally turn verbose JS filesystem code into elegant one-liners
www.npmjs.com/package/@sy... NPM package.
A library to finally turn verbose JS filesystem code into elegant one-liners
July 28, 2025 at 11:00 AM
If you are tired as me of working with native path and fs libraries in NodeJS, you might like my
www.npmjs.com/package/@sy... NPM package.
A library to finally turn verbose JS filesystem code into elegant one-liners
www.npmjs.com/package/@sy... NPM package.
A library to finally turn verbose JS filesystem code into elegant one-liners
Cursor has a system prompt issue.
I'm using LLMs for a complex software migrations and have leveraged everything I know about prompting.
While the model is the same as Claude Code, Cursor seems unable to follow the provided instructions.
Cursor simply feels broken now
I'm using LLMs for a complex software migrations and have leveraged everything I know about prompting.
While the model is the same as Claude Code, Cursor seems unable to follow the provided instructions.
Cursor simply feels broken now
July 25, 2025 at 10:56 AM
Cursor has a system prompt issue.
I'm using LLMs for a complex software migrations and have leveraged everything I know about prompting.
While the model is the same as Claude Code, Cursor seems unable to follow the provided instructions.
Cursor simply feels broken now
I'm using LLMs for a complex software migrations and have leveraged everything I know about prompting.
While the model is the same as Claude Code, Cursor seems unable to follow the provided instructions.
Cursor simply feels broken now
LLMs have helped me build dozens of libraries I've always wanted in Javascript
My favorite one is
www.npmjs.com/package/@sy...
The concept is simple, what if you could write clean template strings in JS ?
I've been using it for clean error formating, LLM messages, and more
My favorite one is
www.npmjs.com/package/@sy...
The concept is simple, what if you could write clean template strings in JS ?
I've been using it for clean error formating, LLM messages, and more
July 23, 2025 at 11:00 AM
LLMs have helped me build dozens of libraries I've always wanted in Javascript
My favorite one is
www.npmjs.com/package/@sy...
The concept is simple, what if you could write clean template strings in JS ?
I've been using it for clean error formating, LLM messages, and more
My favorite one is
www.npmjs.com/package/@sy...
The concept is simple, what if you could write clean template strings in JS ?
I've been using it for clean error formating, LLM messages, and more
When I'm hired to enable a team with AI tools what I'm actually paid for is asking the right questions to reverse engineer how the team works
When I have that, building the tools is the easy part
Also why I fear that the agents trend will do a lot of harm to craftsmanship
When I have that, building the tools is the easy part
Also why I fear that the agents trend will do a lot of harm to craftsmanship
April 7, 2025 at 11:00 AM
When I'm hired to enable a team with AI tools what I'm actually paid for is asking the right questions to reverse engineer how the team works
When I have that, building the tools is the easy part
Also why I fear that the agents trend will do a lot of harm to craftsmanship
When I have that, building the tools is the easy part
Also why I fear that the agents trend will do a lot of harm to craftsmanship
If you're building LLM based products, check out napkin.ai
Blending pre-made templates with LLM formatting may seem like a straightforward idea but they've executed it really well with a clean UX
And if you need to fill a presentation with corporate BS, just ask it to generate the text as well 😅
Blending pre-made templates with LLM formatting may seem like a straightforward idea but they've executed it really well with a clean UX
And if you need to fill a presentation with corporate BS, just ask it to generate the text as well 😅
March 31, 2025 at 3:06 PM
If you're building LLM based products, check out napkin.ai
Blending pre-made templates with LLM formatting may seem like a straightforward idea but they've executed it really well with a clean UX
And if you need to fill a presentation with corporate BS, just ask it to generate the text as well 😅
Blending pre-made templates with LLM formatting may seem like a straightforward idea but they've executed it really well with a clean UX
And if you need to fill a presentation with corporate BS, just ask it to generate the text as well 😅
I've used vibe coding (AI codes without you reviewing) to build a visualizer for a graph with Cline
Would never trust a core product or mission critical system coded like this
People are mistaking vibe coding and vibe engineering
youtube.com/shorts/1_rSr...
Would never trust a core product or mission critical system coded like this
People are mistaking vibe coding and vibe engineering
youtube.com/shorts/1_rSr...
Vibe Coding with AI in 2025
YouTube video by Programmers are also human
youtube.com
March 18, 2025 at 8:16 AM
I've used vibe coding (AI codes without you reviewing) to build a visualizer for a graph with Cline
Would never trust a core product or mission critical system coded like this
People are mistaking vibe coding and vibe engineering
youtube.com/shorts/1_rSr...
Would never trust a core product or mission critical system coded like this
People are mistaking vibe coding and vibe engineering
youtube.com/shorts/1_rSr...
An upside of LLMs is that, just by using them, people will get better at providing instructions and the right context
The downside is that they'll likely stay to get frustrated when you take over 1mn to respond
The downside is that they'll likely stay to get frustrated when you take over 1mn to respond
March 7, 2025 at 11:16 PM
An upside of LLMs is that, just by using them, people will get better at providing instructions and the right context
The downside is that they'll likely stay to get frustrated when you take over 1mn to respond
The downside is that they'll likely stay to get frustrated when you take over 1mn to respond
While illustrating the many disciplines connected to AI and LLMs I realized that I don't have a term for UX Designers / PMs focused on AI
Really shows how under-explored UX for AI is
I agree it's hard, but we could be doing so much more than chats
Really shows how under-explored UX for AI is
I agree it's hard, but we could be doing so much more than chats
March 6, 2025 at 1:04 PM
While illustrating the many disciplines connected to AI and LLMs I realized that I don't have a term for UX Designers / PMs focused on AI
Really shows how under-explored UX for AI is
I agree it's hard, but we could be doing so much more than chats
Really shows how under-explored UX for AI is
I agree it's hard, but we could be doing so much more than chats
Has anyone studied the impact of image and sound on multi-modal LLMs ?
Like if you're working with the LLM, does passing it classical music unrelated to the task influences its response ? What about art pieces ?
Like if you're working with the LLM, does passing it classical music unrelated to the task influences its response ? What about art pieces ?
March 4, 2025 at 12:01 PM
Has anyone studied the impact of image and sound on multi-modal LLMs ?
Like if you're working with the LLM, does passing it classical music unrelated to the task influences its response ? What about art pieces ?
Like if you're working with the LLM, does passing it classical music unrelated to the task influences its response ? What about art pieces ?
Wow, @harper.lol's workflow is incredibly efficient with gemini 2 flash thinking and its massive context
I used to build custom backends to iterate on ideas step by step. Now I just need the LLM
We're truly shifting from coding LLMs to engineering ones
harper.blog/2025/02/16/m...
I used to build custom backends to iterate on ideas step by step. Now I just need the LLM
We're truly shifting from coding LLMs to engineering ones
harper.blog/2025/02/16/m...
My LLM codegen workflow atm
A detailed walkthrough of my current workflow for using LLms to build software, from brainstorming through planning and execution.
harper.blog
February 20, 2025 at 11:42 AM
Wow, @harper.lol's workflow is incredibly efficient with gemini 2 flash thinking and its massive context
I used to build custom backends to iterate on ideas step by step. Now I just need the LLM
We're truly shifting from coding LLMs to engineering ones
harper.blog/2025/02/16/m...
I used to build custom backends to iterate on ideas step by step. Now I just need the LLM
We're truly shifting from coding LLMs to engineering ones
harper.blog/2025/02/16/m...
Finally! Our React Native London talk with @fab-ber.bsky.social on AI assisted migrations is available online
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
Fabrice Bernhard & Yann Leflour - Swift to React Native: AI-Driven Transpilation | RNLConf 2024
YouTube video by Theodo UK
www.youtube.com
February 18, 2025 at 4:14 PM
Finally! Our React Native London talk with @fab-ber.bsky.social on AI assisted migrations is available online
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
Discover how we achieved a 9 months migration in just 4 months by combining solid engineering expertise with a custom AI workflow
www.youtube.com/watch?v=sz_S...
This also means that Deepseek R1 will be a potential attack vector. Willingly put inside a state system
Imagine if they trained the model to act maliciously when it figures out that it's being used inside a targeted infrastructure
I wonder what it could realistically achieve
Imagine if they trained the model to act maliciously when it figures out that it's being used inside a targeted infrastructure
I wonder what it could realistically achieve
India lauds Chinese AI lab DeepSeek, plans to host its models on local servers.
The approval to allow DeepSeek to be hosted in India appears contingent on the platform storing and processing all Indian users’ data domestically, in line with India’s strict data localization requirements.
The approval to allow DeepSeek to be hosted in India appears contingent on the platform storing and processing all Indian users’ data domestically, in line with India’s strict data localization requirements.
January 30, 2025 at 12:05 PM
This also means that Deepseek R1 will be a potential attack vector. Willingly put inside a state system
Imagine if they trained the model to act maliciously when it figures out that it's being used inside a targeted infrastructure
I wonder what it could realistically achieve
Imagine if they trained the model to act maliciously when it figures out that it's being used inside a targeted infrastructure
I wonder what it could realistically achieve
Reposted by Yann Leflour
you can now secretly sign in at tldraw.com/preview
January 28, 2025 at 5:48 PM
you can now secretly sign in at tldraw.com/preview
"Gemini 2 Flash Thinking" is a very impressive model
I'm having discussions with it about high level and abstract concepts and it just "gets it"
The 1M tokens context also means that you can have extremely long conversations with high retrieval rates and without running out of chat history
I'm having discussions with it about high level and abstract concepts and it just "gets it"
The 1M tokens context also means that you can have extremely long conversations with high retrieval rates and without running out of chat history
January 23, 2025 at 10:38 PM
"Gemini 2 Flash Thinking" is a very impressive model
I'm having discussions with it about high level and abstract concepts and it just "gets it"
The 1M tokens context also means that you can have extremely long conversations with high retrieval rates and without running out of chat history
I'm having discussions with it about high level and abstract concepts and it just "gets it"
The 1M tokens context also means that you can have extremely long conversations with high retrieval rates and without running out of chat history
Quick PSA to every developer applying for their first job:
It's time to include LLM tools in your resume (Copilot, Cursor, ...) along with any expirement you've done with them
I'm currently reviewing resumes and none mention these, so it's still a good way to set your profile appart
It's time to include LLM tools in your resume (Copilot, Cursor, ...) along with any expirement you've done with them
I'm currently reviewing resumes and none mention these, so it's still a good way to set your profile appart
January 21, 2025 at 3:37 PM
Quick PSA to every developer applying for their first job:
It's time to include LLM tools in your resume (Copilot, Cursor, ...) along with any expirement you've done with them
I'm currently reviewing resumes and none mention these, so it's still a good way to set your profile appart
It's time to include LLM tools in your resume (Copilot, Cursor, ...) along with any expirement you've done with them
I'm currently reviewing resumes and none mention these, so it's still a good way to set your profile appart
I stopped my #devin (yes the "AI Engineer") subscription, felt too uneasy about merging those pull requests
It was a fun (though way too expensive) trial
It's hard to put it, but to me it's like most "AI art", the demo is impressive but it clearly lacks any intent
It was a fun (though way too expensive) trial
It's hard to put it, but to me it's like most "AI art", the demo is impressive but it clearly lacks any intent
January 7, 2025 at 12:00 PM
I stopped my #devin (yes the "AI Engineer") subscription, felt too uneasy about merging those pull requests
It was a fun (though way too expensive) trial
It's hard to put it, but to me it's like most "AI art", the demo is impressive but it clearly lacks any intent
It was a fun (though way too expensive) trial
It's hard to put it, but to me it's like most "AI art", the demo is impressive but it clearly lacks any intent
@vercel.com's AI SDK is amazing to work with
But I don't get why I had to implement a middleware to add the last message from the prompt to the result if its role is "assistant"
Should be a default behavior
Anyway, if you need it it's "includeAssistantMessage" in @synstack/llm/middleware
But I don't get why I had to implement a middleware to add the last message from the prompt to the result if its role is "assistant"
Should be a default behavior
Anyway, if you need it it's "includeAssistantMessage" in @synstack/llm/middleware
January 6, 2025 at 10:11 PM
@vercel.com's AI SDK is amazing to work with
But I don't get why I had to implement a middleware to add the last message from the prompt to the result if its role is "assistant"
Should be a default behavior
Anyway, if you need it it's "includeAssistantMessage" in @synstack/llm/middleware
But I don't get why I had to implement a middleware to add the last message from the prompt to the result if its role is "assistant"
Should be a default behavior
Anyway, if you need it it's "includeAssistantMessage" in @synstack/llm/middleware
Even though I would not be surprised to see something like this implemented, there is no way it's true
But seeing how this is spreading, it shows that we're far from achieving widespread understanding of what AI really is
But seeing how this is spreading, it shows that we're far from achieving widespread understanding of what AI really is
“1984” thrives at Twitter; this is the directive from Mr Absolute Free Speech:
January 6, 2025 at 6:17 PM
Even though I would not be surprised to see something like this implemented, there is no way it's true
But seeing how this is spreading, it shows that we're far from achieving widespread understanding of what AI really is
But seeing how this is spreading, it shows that we're far from achieving widespread understanding of what AI really is
Finally released @synstack/synscript v2 with the #llm package migrated to @vercel.com's AI SDK 🎉
The Llm library is now a chainable, immutable, and type-safe wrapper over the ai sdk
If you want to improve your prompting DevX in Typescript, look no further
github.com/pAIrprogio/...
The Llm library is now a chainable, immutable, and type-safe wrapper over the ai sdk
If you want to improve your prompting DevX in Typescript, look no further
github.com/pAIrprogio/...
synscript/packages/llm at main · pAIrprogio/synscript
Libraries for immutable and chainable LLM compatible scripting - pAIrprogio/synscript
github.com
January 6, 2025 at 12:45 PM
Finally released @synstack/synscript v2 with the #llm package migrated to @vercel.com's AI SDK 🎉
The Llm library is now a chainable, immutable, and type-safe wrapper over the ai sdk
If you want to improve your prompting DevX in Typescript, look no further
github.com/pAIrprogio/...
The Llm library is now a chainable, immutable, and type-safe wrapper over the ai sdk
If you want to improve your prompting DevX in Typescript, look no further
github.com/pAIrprogio/...
I got a nifty TS challenge for you @mattpocock.com
It's for chained flat object merging and there's no reccursive generics in play
Any way to prevent "Type instantiation is excessively deep and possibly infinite.(2589)" on this one ?
www.typescriptlang.org/play/#code/C...
It's for chained flat object merging and there's no reccursive generics in play
Any way to prevent "Type instantiation is excessively deep and possibly infinite.(2589)" on this one ?
www.typescriptlang.org/play/#code/C...
TS Playground - An online editor for exploring TypeScript and JavaScript
The Playground lets you write TypeScript or JavaScript online in a safe and sharable way.
www.typescriptlang.org
December 31, 2024 at 5:02 PM
I got a nifty TS challenge for you @mattpocock.com
It's for chained flat object merging and there's no reccursive generics in play
Any way to prevent "Type instantiation is excessively deep and possibly infinite.(2589)" on this one ?
www.typescriptlang.org/play/#code/C...
It's for chained flat object merging and there's no reccursive generics in play
Any way to prevent "Type instantiation is excessively deep and possibly infinite.(2589)" on this one ?
www.typescriptlang.org/play/#code/C...
Finally tried Evalite, the TS eval library for #LLM from @mattpocock.com
github.com/mattpocock/e...
So far I really like it:
- Advises for external libraries instead of reinventing the wheel
- Pure TS as expected
- Integrates into your project rather than constrain it
github.com/mattpocock/e...
So far I really like it:
- Advises for external libraries instead of reinventing the wheel
- Pure TS as expected
- Integrates into your project rather than constrain it
GitHub - mattpocock/evalite: Test your LLM-powered apps with TypeScript. No API key required.
Test your LLM-powered apps with TypeScript. No API key required. - mattpocock/evalite
github.com
December 29, 2024 at 7:59 PM
Finally tried Evalite, the TS eval library for #LLM from @mattpocock.com
github.com/mattpocock/e...
So far I really like it:
- Advises for external libraries instead of reinventing the wheel
- Pure TS as expected
- Integrates into your project rather than constrain it
github.com/mattpocock/e...
So far I really like it:
- Advises for external libraries instead of reinventing the wheel
- Pure TS as expected
- Integrates into your project rather than constrain it
This doc pull request by #Devin is rotting in my backlog
github.com/pAIrprogio/s...
While I "kind of" agree with what Devin wrote, I trully feel uneasy about merging it
What bugs me most is that the LLM is paraphrasing what I wrote manually for other packages without any clear intent
github.com/pAIrprogio/s...
While I "kind of" agree with what Devin wrote, I trully feel uneasy about merging it
What bugs me most is that the LLM is paraphrasing what I wrote manually for other packages without any clear intent
docs: add comprehensive documentation for all packages by devin-ai-integration[bot] · Pull Request #5 · pAIrprogio/synscript
Documentation Update
Changes per Package
enhance: Update documentation to clarify object-only usage and remove primitive type examples
llm: Restructure documentation into separate files and emphas...
github.com
December 20, 2024 at 2:06 PM
This doc pull request by #Devin is rotting in my backlog
github.com/pAIrprogio/s...
While I "kind of" agree with what Devin wrote, I trully feel uneasy about merging it
What bugs me most is that the LLM is paraphrasing what I wrote manually for other packages without any clear intent
github.com/pAIrprogio/s...
While I "kind of" agree with what Devin wrote, I trully feel uneasy about merging it
What bugs me most is that the LLM is paraphrasing what I wrote manually for other packages without any clear intent