fix(opencode): add input limit for compaction - #8465

Hey! Your PR title openai: add input limit for codex-5.2 doesn't follow conventional commit format.

Please update it to start with one of:

• feat: or feat(scope): new feature
• fix: or fix(scope): bug fix
• docs: or docs(scope): documentation changes
• chore: or chore(scope): maintenance tasks
• refactor: or refactor(scope): code refactoring
• test: or test(scope): adding or updating tests

Where scope is the package name (e.g., app, desktop, opencode).

See CONTRIBUTING.md for details.

The following comment was made by an LLM, it may be inaccurate:

No duplicate PRs found

Auto-compaction now triggers based on the optional input limit.

There is a separate issue (I believe with all models) where the next outgoing message can overrun context/input limits. In this scenario, the error should be caught with an auto-compaction running on thread history, and then a retry of the latest message occurring. Perhaps integrated into SessionRetry.retryable with a force auto-compaction. I won't attempt that in this PR.

/review

What a G

lgtm

Auto-compaction now triggers based on the optional input limit.

There is a separate issue (I believe with all models) where the next outgoing message can overrun context/input limits. In this scenario, the error should be caught with an auto-compaction running on thread history, and then a retry of the latest message occurring. Perhaps integrated into SessionRetry.retryable with a force auto-compaction. I won't attempt that in this PR.

Yeah i mean this is awesome so thank you! Interesting idea would like to see it

Can this be an option to deal with models that get a bit weird or just plain broken long before their context window ends? The zai coding plan really starts to degrade fast into errors for a lot of us (See other threads) around 120k context. It would be interesting to be able to set this limit to 120k and have it auto compact at that point for that model.

@greggh the input limit is available in the config. You could provide a custom model config:

"providerId": {
    "models": {
      "modelId": {
        "limit": { "input": 120000, "context": 200000 },
      },
    },
  },