3 ms·

> Every API call sends the full tool schema for all available tools. Only if you are doing it wrong, search >>> summarization Then the other question, is it d

> Every API call sends the full tool schema for all available tools. Only if you are doing it wrong, search >>> summarization Then the other question, is it deterministic between runs or am I going to get a different summary each session, turn, or toolcall? And depending on that frequency, am I using more token than I save by doing summarization for N tools? Minimizing token usage is not the goal in of itself, re: the ageless tradeoff of quantity vs quality For some context, my system prompt is around 5k tokens at the start. I put file contents there read/write/agents.md, which save millions of tokens and seems to work better than making them message parts. > Just trimmed boilerplate This is not what I see this tool doing. It's automatically manipulating words in the background that you should put far more care and attention towards. Referring to those words as "boilerplate" you can just throw into a slop machines is revealing

aSidorenkoCode 9mo ago
the benchmarks show no degradation in task completion with the shorter descriptions. We're in the age where frontier LLMs don't need instructions on how to read or edit a file. The descriptions aren't dynamically summarized either. They're static in the plugin, same every call, every session. Zero overhead, fully deterministic. This has been validated in over 3000 benchmark runs in OpenCode and I ran the entire Exercism Python practice suite (https://github.com/exercism/python/tree/main/exercises/practice https://github.com/exercism/python/tree/main/exercises/pract...) with and without the plugin with identical results. An initial dataset is shared in the repo.
- verdverm 9mo ago
  Have you made that benchmarking process open so others could reproduce it? > with identical results If your results are identical, you should be very sus, something is wrong if this is true. Nothing in agentic is reliable of fully deterministic
  - aSidorenkoCode 9mo ago
    Good benchmark results don't mean identical outputs. The task completion rate is the same: both pass the same exercises. The paths the model takes differ, but the end result is the same -> pass the tests The full benchmarking methodology and tooling will be published alongside the paper.
    - verdverm 9mo ago
      you used the word "identical" to describe it, not me words matter which is why I still think this is a terrible idea, I don't think it holds up in the general case and would, as a peer reviewer, be inclined to believe there is benchmark filtering that makes for good results. You should use the same benchmarks everyone else is when you write your paper
      - verdverm 9mo ago
        timely result, https://arxiv.org/pdf/2602.12670 https://arxiv.org/pdf/2602.12670

aSidorenkoCode 9mo ago

the benchmarks show no degradation in task completion with the shorter descriptions. We're in the age where frontier LLMs don't need instructions on how to read or edit a file. The descriptions aren't dynamically summarized either. They're static in the plugin, same every call, every session. Zero overhead, fully deterministic. This has been validated in over 3000 benchmark runs in OpenCode and I ran the entire Exercism Python practice suite (https://github.com/exercism/python/tree/main/exercises/practice https://github.com/exercism/python/tree/main/exercises/pract...) with and without the plugin with identical results. An initial dataset is shared in the repo.

verdverm 9mo ago
Have you made that benchmarking process open so others could reproduce it? > with identical results If your results are identical, you should be very sus, something is wrong if this is true. Nothing in agentic is reliable of fully deterministic
- aSidorenkoCode 9mo ago
  Good benchmark results don't mean identical outputs. The task completion rate is the same: both pass the same exercises. The paths the model takes differ, but the end result is the same -> pass the tests The full benchmarking methodology and tooling will be published alongside the paper.
  - verdverm 9mo ago
    you used the word "identical" to describe it, not me words matter which is why I still think this is a terrible idea, I don't think it holds up in the general case and would, as a peer reviewer, be inclined to believe there is benchmark filtering that makes for good results. You should use the same benchmarks everyone else is when you write your paper
    - verdverm 9mo ago
      timely result, https://arxiv.org/pdf/2602.12670 https://arxiv.org/pdf/2602.12670