[flagged]
How to code Claude Code in 200 lines of code
21–30 of 249 posts
Re: How to code Claude Code in 200 lines of code
#22yes, it's an agent
Re: How to code Claude Code in 200 lines of code
#23This reflects my experience. Yet, I feel that getting reliability out of LLM calls with a while-loop harness is elusive. For example - how can I reliably have a decision block to end the loop (or keep it running)? - how can I reliably call tools with the right schema? - how can I reliably summarize context / excise noise from the conversation? Perhaps, as the models get better, they'll approach some threshold where m…
Re (1) use a TODOs system like Claude code.
Re (2) also fairly easy! It's just a summarization prompt. E.g. this is the one we use in our agent: https://github.com/HolmesGPT/holmesgpt/blob/62c3898e4efae69b...
Or just use the Claude Code SDK that does this all for you! (You can also use various provider-specific features for 2 like automatic compaction on OpenAI responses endpoint.)
Re: How to code Claude Code in 200 lines of code
#24The tip of the sphere in agentic code harnesses today is to RL train them as dedicated conductor/orchestrator models.
Not 200 lines of Python.
Re: How to code Claude Code in 200 lines of code
#25This reflects my experience. Yet, I feel that getting reliability out of LLM calls with a while-loop harness is elusive. For example - how can I reliably have a decision block to end the loop (or keep it running)? - how can I reliably call tools with the right schema? - how can I reliably summarize context / excise noise from the conversation? Perhaps, as the models get better, they'll approach some threshold where m…
Forgot to address the easiest part:
> - how can I reliably call tools with the right schema?
This is typically done by enabling strict mode for tool calling which is a hermetic solution. Makes llm unable to generate tokens that would violate the schema. (I.e. LLM samples tokens only from the subset of tokens that lead to valid schema generation.)
Re: How to code Claude Code in 200 lines of code
#26[flagged]
[flagged]
Re: How to code Claude Code in 200 lines of code
#27Re: How to code Claude Code in 200 lines of code
#28Earlier quoted context omitted.
But does that extra complexity actually improve performance? https://www.tbench.ai/leaderboard/terminal-bench/2.0 says yes, but not as much as you'd think. "Terminus" is basically just a tmux session and LLM in a loop.
I'm not a good representative for claude code because I'm primarily a codex user now, but I know that if codex had subagents it would be at least twice as productive. Time spent is an important aspect of performance so yup, the complexity improved performance.
Not necessarily true. Subagents allow for parallelization but they can decrease accuracy dramatically if you're not careful because there are often dependencies between tasks and swapping context windows with a summary is extremely lossy.
For the longest time, Claude Code itself didnt really use subagents much by default, other than supporting them as a feature eager users could configure. (Source is reverse engineering we did on Claude code using the fantastic CC tracing tool Simon Willison wrote about once. This is also no longer true on latest versions that have e.g. an Explore subagent that is actively used.)
Re: How to code Claude Code in 200 lines of code
#29Earlier quoted context omitted.
But does that extra complexity actually improve performance? https://www.tbench.ai/leaderboard/terminal-bench/2.0 says yes, but not as much as you'd think. "Terminus" is basically just a tmux session and LLM in a loop.
I'm not a good representative for claude code because I'm primarily a codex user now, but I know that if codex had subagents it would be at least twice as productive. Time spent is an important aspect of performance so yup, the complexity improved performance.
Are subagents a fundamental change, or just acting as inner loops to the agentic loop similar to the one in the article?
Re: How to code Claude Code in 200 lines of code
#30This article was more true than not a year ago but now the harnesses are so far past the simple agent loop that I'd argue that this is not even close to an accurate mental model of what claude code is doing.
Obviously modern harnesses have better features but I wouldn't say it invalidates the mental model. Simpler agents aren't that far behind in performance if the underlying model is the same, including very minimal ones with basic tools. I'd say it's similar to how a "make your own relational DB" article might feature a basic B-tree with merge-joins. Yeah, obviously real engines have sophisticated planners, multiple jo…
You’re not wrong but I still think that the harness matters a lot when trying to accurately describe Claude Code.
Here’s a reframing:
If you asked people “what would you rather work with, today’s Claude Code harness with sonnet 3.7, or the 200 line agentic loop in the article with Opus 4.5, which would you choose?”
I suspect many people would choose 3.7 with the harness. Moreover, that is true, then I’d say the article is no longer useful for a modern understanding of Claude Code.