Skip to content

Codex mini - #34

Merged
adamdotdevin merged 3 commits into
anomalyco:devfrom
PhantomReactor:codex-mini
May 18, 2025
Merged

Codex mini#34
adamdotdevin merged 3 commits into
anomalyco:devfrom
PhantomReactor:codex-mini

Conversation

@PhantomReactor

@PhantomReactorPhantomReactor commented May 18, 2025

Copy link
Copy Markdown
Contributor

Closes#27

@adamdotdevin
adamdotdevin merged commit 2f8984f into anomalyco:devMay 18, 2025
@adamdotdevin

Copy link
Copy Markdown
Member

nice lol

randomm referenced this pull request in randomm/opencode Jan 9, 2026
…#34)
Previous patterns used incorrect glob syntax that didn't match
AWS CLI command format:
- Wrong: aws */describe* (expected slash)
- Wrong: aws *:create* (expected colon)
- Actual AWS CLI: aws ec2 describe-instances (space-separated)
Fixed all 30+ patterns to use correct format:
- Correct: aws * describe* (space wildcard space action)
- Matches: aws <service> <action> format
Changes applied to both personal and work configs.
Enables research-specialist to actually query AWS services.
xywsxp pushed a commit to xywsxp/opencode that referenced this pull request Apr 24, 2026
"github.com/openai/openai-go/shared"
"github.com/sst/opencode/internal/config"
"github.com/sst/opencode/internal/status"
)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggestion (import ordering): Standard library imports should come first, then external packages, then internal packages. Consider reordering imports like this:

import (
"context""encoding/json""errors""fmt""io""log/slog""time""github.com/openai/openai-go""github.com/openai/openai-go/responses""github.com/openai/openai-go/shared""github.com/sst/opencode/internal/config""github.com/sst/opencode/internal/llm/models""github.com/sst/opencode/internal/llm/tools""github.com/sst/opencode/internal/message""github.com/sst/opencode/internal/status"
)

This is a minor suggestion for consistency with the style guide.

Messages: messages,
Tools: tools,
}
if o.providerOptions.model.CanReason == true {

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggestion (style): Using == true is redundant in Go. Consider simplifying to just if o.providerOptions.model.CanReason {. This is a minor nitpick for Go idiomatic style.


params.MaxOutputTokens = openai.Int(o.providerOptions.maxTokens)

if o.providerOptions.model.CanReason == true {

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggestion (style): Same here - using == true is redundant in Go. Consider simplifying to just if o.providerOptions.model.CanReason {.


return outputTools
}

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggestion (formatting): There are multiple consecutive blank lines here. Running go fmt should fix this - standard Go formatting uses single blank lines for separation.

@rootxharsh

Copy link
Copy Markdown

Code Review Summary

I've reviewed the PR changes (note: the actual changes are related to OpenAI provider refactoring and adding Codex Mini support, not TUI rendering as mentioned in the description).

Minor Suggestions (all optional):

  1. Import ordering in openai_response.go - Standard library imports should come first, then external packages, then internal packages per the style guide.

  2. Boolean comparison style - Using == true is redundant in Go. Found in:

    • openai_completion.go:108
    • openai_response.go:135
  3. Formatting - Multiple consecutive blank lines in a few places. Running go fmt should clean these up.

Overall

The code logic looks solid. The refactoring to split chat completion and response APIs into separate files is a good approach for maintainability. The new Codex Mini model integration follows the existing patterns well.

These are all minor suggestions - feel free to address them or leave as-is based on your preference.

bussard76 pushed a commit to bussard76/openwork that referenced this pull request May 12, 2026
avion23 pushed a commit to avion23/opencode that referenced this pull request Jun 10, 2026
zoulukuang added a commit to zoulukuang/deskfox that referenced this pull request Aug 14, 2026
- anomalyco#34 复制:`clipboard.writeText` **依赖 document 有焦点**,焦点被抢时静默失败 ——
表现为偶发「点了复制但剪贴板没变」。加 `Page.bringToFront` + 失败重试一次。
(此前两条弯路都记在注释里:clipboard API 读不了、CDP 合成 ⌘V 触发不了粘贴。)
- anomalyco#37 步骤展开/收起:会话变长后步骤被虚拟列表回收,视口内一个都没有 ——
这是**前提没准备好**,不是没有步骤。改为先跳到最新;仍没有就现跑一条 shell 命令造一个。
直接记 SKIP 等于会话一长这条就永远验不到。
- anomalyco#30 撤销/重做:撤销是异步的,改为等消息数真的变而不是固定 sleep(读早了必然误判)。
第 3 组现状:通过 15,跳过 1(分享 —— 需 user 逐次授权,已在授权下单独验过:
生成链接 → 取消分享 → curl 实测失效)。
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Wut6R6htBzR8fBN8gbJkWg
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

codex-mini, openai responses api

3 participants

@PhantomReactor@adamdotdevin@rootxharsh