back
user profile

tosh

183,935karma·27,647submissions·May 4, 2010
about
https://smolenv.com (a smol agent)

https://findableapp.com (SEO toolkit for Google & ChatGPT)

https://kiwilang.com (k-like language and implementation in Zig with support for GPU via Apple MLX)

https://chesscatsapp.com (a fun way to play chess)

https://moreepisodes.com (tv show episodes generated by GPT-4)

https://jamshelf.com (open source Clubhouse)

https://magic.do (early stage fund)

https://lemmings.io (sci-fi themed hackathons (think "zombie apocalypse", "aliens"))

https://applesilicongames.com (game compatibility and game performance on Apple Silicon Macs)

https://devmonthly.com (curated news & input & jobs for software engineers)

https://blossom.io (project tracking for distributed teams)

more from me around the web:

https://twitter.com/__tosh

https://github.com/tosh

https://angel.co/tosh

https://medium.com/@__tosh

https://lobste.rs/u/tosh

https://dribbble.com/tosh

https://linkedin.com/in/tschranz

https://instagram.com/thomas.schranz

https://facebook.com/thomas.schranz

{UX Service Game} Design, Typography, Clojure, Lisp, Python, Tea, Minimalism, Dao, …

[ my public key: https://keybase.io/tosh; my proof: https://keybase.io/tosh/sigs/gBG56O339IbBPh1iaudGL3dgJuX2yONIATYv62xHXdg ]

recent activity (27,647 total)
comment
your favorite agent will help you de-golf and analyze the implementation I'm confident you can understand what it does (and does not do) and why in an afternoon (probably in 20-30 minutes actuall…
11d ago·view thread
comment
> feeds everything into bash, making it completely pointless here is a typical bench run with 9 runs and traces for opencode, pi and smol you can look at every step and which tools are used and how…
11d ago·view thread
comment
i guess what i'm saying is: it's not so hard to start with a minimal, well working agent and then only add exactly what you need instead of taking a more complex agent and then trimming it d…
11d ago·view thread
comment
smol is ~20 lines of Go, uses only stdlib but also highly opinionated (no mcps, no agents.md, no system prompt, …) so not sure it checks all of your boxes that said: because smol is so smol you can ad…
11d ago·view thread
comment
it's very difficult and time-consuming to make good comparisons (especially for open ended tasks) also because there are so many configuration options - which env do you provide? - which mo…
11d ago·view thread
comment
great point, progressive disclosure is a good pattern for context management
11d ago·view thread
comment
I'm glad there are quite a few good open source ones by now. Also happy with how much love codex gets from OpenAI. That said: I was looking at existing agents to find one to build upon and to me …
11d ago·view thread
comment
I'm glad MCP is getting simpler a few months ago I tried to implement an MCP server from scratch in python (instead of using the existing reference implementation) and I could not get it to work …
11d ago·view thread
comment
a few examples: 1) system prompt in pi is quite small (way smaller than the one from OpenCode) 2) when your agents.md file changes pi does not re-spam it (preserves cache, good trade-off!) 3) only 4 t…
11d ago·view thread
comment
I use it for smaller changes (no compaction yet) If I have agents.md or other context I want it to read I mention it at the beginning of the session re MCP: I am not using an MCP with smol but there a…
11d ago·view thread
comment
i also found re context window: less is more at least with GPT 5.6 Sol fwiw https://smolenv.com/t/nested-template-includes-60636/ sh is all you need…
11d ago·view thread
comment
in this case there was a hidden grader that checked if the implementation was correct (because that was the easiest thing to check), all 3 agents cleared this hurdle in all 9 runs I agree, next it mak…
12d ago·view thread
comment
apologies, I should have clarified the 'better' claim - same task result (passed) - finished faster - fewer tokens, less cost - fewer requests for inference - fewer tool calls …
12d ago·view thread
comment
maybe a bit counter-intuitive but: I found that removing - system prompt - skills - agents.md - mcps + reducing tools to just 1 (sh) gives better results than having 'more' of t…
12d ago·view thread
comment
I agree in principle but in practice this is not easy even with a robust, well thought through setup every model behaves in its own way, some adhere more to a system prompt, another model has more rec…
12d ago·view thread
comment
Pi is pretty good actually (it has way less context spam, fewer tools and smaller system prompt than the usual suspects) databricks also looked at this: https://earendil.com/posts/…
12d ago·view thread
comment
I'm not saying system prompts, agents.md, tools, skills don't have their place actually the opposite: they matter a lot because they do steer the agent with great power comes great responsib…
12d ago·view thread
comment
I know it sounds a bit counter-intuitive but try a fresh coding session without skills, agents.md, system prompt and additional tools I think you will be positively surprised how good current models l…
12d ago·view thread