[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80198":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":15,"stars7d":17,"stars30d":18,"stars90d":16,"forks30d":16,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":22,"hasPages":22,"topics":24,"createdAt":10,"pushedAt":10,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":16,"starSnapshotCount":16,"syncStatus":17,"lastSyncTime":28,"discoverSource":29},80198,"CodeSeeX","TasteSteak\u002FCodeSeeX","TasteSteak","CodeSeeX is a local desktop bridge that lets Codex run DeepSeek V4 with 1M context, Codex tool compatibility, built-in Web Search, and a ready-to-copy config.toml setup.","",null,"Rust",83,7,76,1,0,2,6,3,2.71,"GNU Affero General Public License v3.0",false,"main",[],"2026-06-12 02:03:59","\u003Ch1 align=\"center\">CodeSeeX\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n  \u003Cimg alt=\"Version 0.4.0\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fversion-0.4.0-1f6feb\">\n  \u003Cimg alt=\"Platform Windows macOS Linux\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fplatform-Windows%20%7C%20macOS%20%7C%20Linux-2ea043\">\n  \u003Cimg alt=\"License AGPL-3.0-only\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-AGPL--3.0--only-bd561d\">\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  Run DeepSeek V4 in Codex with 1M context, Codex tool compatibility, and built-in Web Search.\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  Unofficial and unaffiliated. Use your own credentials and follow the applicable Codex, OpenAI, DeepSeek, and search-provider terms.\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cimg alt=\"CodeSeeX dashboard showing service status and balance\" src=\"docs\u002Fimg\u002Fdashboard-balance.png\">\n\u003C\u002Fp>\n\nCodeSeeX gives Codex a local Responses API endpoint, forwards supported requests to DeepSeek, preserves the Codex tool workflow, and provides a desktop manager for setup, logs, usage, tools, and adapter configuration.\n\n```text\nCodex Desktop  ->  CodeSeeX local API  ->  DeepSeek API\n                       ^\n                       |\n                 desktop manager\n```\n\n## What You Get\n\n- DeepSeek V4 with 1M context: expose `deepseek-v4-pro` and `deepseek-v4-flash` to Codex with million-token catalog metadata.\n- Codex tool compatibility: keep Codex workflows such as Apply Patch, MCP, Skills, Plugins, and native MCP tools available through the bridge.\n- Built-in tool layer: use CodeSeeX Web Search, workspace search, file reading, and patch support out of the box.\n- High-fidelity context state: preserve long `previous_response_id` chains, verified tool facts, and compacted context across multi-turn agent work.\n- Cache-friendly tool conversion: typed screenshot\u002Fimage outputs are represented as stable metadata instead of large base64 text.\n- Reliable catalog bootstrap: ship a full CodeSeeX catalog seed for first-run machines without a native Codex model catalog.\n- Generated setup: copy a ready-to-use `config.toml` from the CodeSeeX proxy settings page.\n- Custom upstream support: point CodeSeeX at the official DeepSeek API or a self-hosted OpenAI-compatible endpoint.\n- Experimental upstream compatibility: keep official DeepSeek `\u002Fv1\u002Fchat\u002Fcompletions` routing enabled by default, with a switch for strict `\u002Fchat\u002Fcompletions` testing.\n- Local visibility: see service state, balance, Flash\u002FPro usage estimates, user-level logs, and tool activity from the desktop UI.\n\n## Quick Start\n\n1. Download the latest build for your platform from [GitHub Releases](https:\u002F\u002Fgithub.com\u002FTasteSteak\u002FCodeSeeX\u002Freleases).\n2. Start CodeSeeX and open the desktop manager.\n3. Go to `Settings -> Proxy`.\n4. Confirm the local service is running. The default listen port is `8787`.\n5. Copy the generated `config.toml` shown by CodeSeeX into the Codex configuration you use for DeepSeek.\n6. Restart CodeSeeX only if you changed the listen port, then start or restart Codex Desktop.\n7. In Codex, choose `deepseek-v4-pro` or `deepseek-v4-flash` and start a new conversation.\n\nRelease artifacts may vary by version. Build scripts are included for Windows, macOS, and Linux, and platform builds should be verified on real devices before release.\n\n## Codex config.toml\n\nPrefer copying the TOML generated inside CodeSeeX, because the local port and catalog path are machine-specific.\n\n```toml\nmodel_provider = \"custom\"\nmodel = \"deepseek-v4-pro\"\ndisable_response_storage = true\nmodel_reasoning_effort = \"xhigh\"\nmodel_catalog_json = '~\u002F.codeseex\u002Fmodel-catalog.json'\n\n[model_providers.custom]\nname = \"DeepSeek\"\nwire_api = \"responses\"\nrequires_openai_auth = true\nbase_url = \"http:\u002F\u002F127.0.0.1:8787\u002Fv1\"\n```\n\nTo use the faster model, change:\n\n```toml\nmodel = \"deepseek-v4-flash\"\n```\n\nCodeSeeX reads API credentials from the user's Codex auth configuration. It does not store DeepSeek API keys in `proxy.env`.\n\nIf you self-host a DeepSeek-compatible service, set the upstream URL in `Settings -> Proxy`. Leave it blank to use the official DeepSeek API.\n\n\u003Cp align=\"center\">\n  \u003Cimg alt=\"CodeSeeX proxy settings with generated config.toml\" src=\"docs\u002Fimg\u002Fconfig-toml.png\">\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cimg alt=\"CodeSeeX usage view showing conversation token and cost details\" src=\"docs\u002Fimg\u002Fusage.png\">\n\u003C\u002Fp>\n\n## Features\n\n- Codex-compatible local API for `\u002Fv1\u002Fresponses` and related model calls.\n- DeepSeek V4 adapter catalog for `deepseek-v4-flash` and `deepseek-v4-pro` with `1M` context metadata.\n- Compatibility with Codex built-in tool flows, including Apply Patch, MCP, Skills, and Plugins.\n- Native MCP passthrough so Codex-configured MCP tools remain executed and displayed by the Codex app tool layer.\n- Native Apply Patch bridge so DeepSeek tool calls are returned to Codex as the built-in freeform patch tool instead of a shell-style wrapper.\n- High-fidelity context compiler with verified tool facts, manual\u002Fautomatic compaction support, crash-safe checkpoints, and safer interrupted-request recovery.\n- Structured tool-output conversion that preserves regular JSON\u002Ftext while omitting large binary payloads before they enter model-visible context.\n- Single configurable local port for the desktop manager, `\u002Fapi\u002F*`, and `\u002Fv1\u002F*`.\n- Proxy settings for catalog mode, upstream model override, custom upstream URL, Flash\u002FPro billing rates, and generated `config.toml`.\n- Streaming answer display with reasoning visibility controls and grouped proxy tool summaries.\n- User-level logs, daily log retention, balance checks, usage estimates, tray shortcuts, auto-start, and silent update indicators.\n- Built-in tools for Web Search, patching, workspace search, and file reading, plus optional community tools.\n\n## Configuration & Tools\n\nRuntime data is stored in the user's `~\u002F.codeseex` directory by default so installed apps can write safely across Windows, macOS, and Linux.\n\nImportant runtime paths:\n\n- `~\u002F.codeseex\u002Fmodel-catalog.json`: adapter catalog referenced by Codex through `model_catalog_json`.\n- `~\u002F.codeseex\u002Fproxy-state.json`: local response chain, compact state, and verified tool facts used to continue Codex conversations.\n- `~\u002F.codeseex\u002Fcompact.key`: local key used for CodeSeeX-readable compact payloads.\n- `~\u002F.codeseex\u002Flogs\u002F`: daily user-level logs such as `logs-20260521.jsonl`.\n- `~\u002F.codeseex\u002Fextension\u002Ftools\u002F\u003Ctool>\u002F`: optional community tool packages.\n- `~\u002F.codeseex\u002Fproxy.env`: non-secret local runtime settings.\n\nCommunity tools are disabled unless enabled in configuration. Enabling community tool code means running local code from that tool package, so only use tools you trust. See [docs\u002Ftool-authoring.md](docs\u002Ftool-authoring.md) for the tool package format.\n\nMCP servers stay on the user's Codex configuration. CodeSeeX translates Codex-provided MCP tool declarations for DeepSeek, then returns native `function_call` items so Codex can execute and display MCP calls itself.\n\nProxy-hosted CodeSeeX tools are separate from MCP. Built-in hosted tools run inside CodeSeeX, and community hosted tools must provide an explicit execution hook; otherwise CodeSeeX returns a clear unsupported-tool result instead of rerouting the call to another tool.\n\nCodeSeeX keeps response state locally and does not auto-delete conversation chains for normal operation. If a request is interrupted, verified user input and tool facts are retained while incomplete assistant text is not reused as final context.\n\n## Troubleshooting\n\n### Balance Query Fails\n\n- Make sure Codex auth is configured for the same user account.\n- Confirm the machine can reach the DeepSeek API endpoint.\n- If a local proxy is configured, make sure that proxy process is actually running.\n\n### Codex Cannot See DeepSeek Models\n\n- Confirm `model_catalog_json` points to an existing `~\u002F.codeseex\u002Fmodel-catalog.json`.\n- Copy the generated TOML from CodeSeeX instead of typing the path manually.\n- Restart Codex after changing TOML.\n- GPT\u002FOpenAI TOML files do not need `model_catalog_json` and are not affected by CodeSeeX.\n\n### Conversation Fails With `fetch failed`\n\n- Check the CodeSeeX logs page for the upstream error.\n- Confirm Codex `base_url` points to CodeSeeX, for example `http:\u002F\u002F127.0.0.1:8787\u002Fv1`.\n- If you use the official upstream, test whether the machine can access `https:\u002F\u002Fapi.deepseek.com`.\n- If you use a self-hosted upstream, confirm the URL is reachable and OpenAI-compatible.\n- For the official DeepSeek upstream, CodeSeeX sends chat requests to the `\u002Fv1\u002Fchat\u002Fcompletions` compatibility endpoint by default. Use `Settings -> Experimental` only if you need to test strict `\u002Fchat\u002Fcompletions` routing.\n- Make sure no other process is using the configured CodeSeeX port.\n\n## Development\n\n```sh\nnpm install\nnpm start\nnpm run check\nnpm run dist:win\n```\n\nAdditional build scripts are available for target-platform testing:\n\n```sh\nnpm run dist:mac\nnpm run dist:linux\n```\n\nGitHub Actions can build the desktop artifacts automatically for Linux, macOS, and Windows. See [Desktop Build Action](docs\u002Fbuild-actions.md) for the workflow setup and testing steps.\n\n## Privacy & License\n\nCodeSeeX is a local proxy, but model requests are forwarded to the configured DeepSeek API endpoint. Do not send code, secrets, personal data, or third-party material unless you have permission to process it with that service.\n\nThe built-in `web_search` tool may request search-result pages or regular web pages from third-party websites. Those services may apply their own terms, rate limits, and anti-abuse rules.\n\nCodeSeeX is licensed under AGPL-3.0-only. See [LICENSE](LICENSE). If you modify and distribute CodeSeeX, or provide a modified version as a network service, you must make the corresponding source code available under the same license.\n","CodeSeeX 是一个本地桌面桥梁，允许Codex运行DeepSeek V4模型，并支持1M上下文长度、Codex工具兼容性以及内置的网络搜索功能。其核心功能包括提供与Codex工具如Apply Patch、MCP等的无缝集成，同时通过内置的Web搜索增强开发效率；此外，它还支持高保真的上下文状态保持和缓存友好的工具转换，确保多轮对话中信息的一致性和准确性。适用于需要在Codex环境中利用更强大语言模型进行开发或研究的场景，尤其是在要求长时间会话记忆和高级搜索能力的情况下。","2026-06-11 03:59:37","CREATED_QUERY"]