[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-4751":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":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":31,"readmeContent":32,"aiSummary":33,"trendingCount":16,"starSnapshotCount":16,"syncStatus":34,"lastSyncTime":35,"discoverSource":36},4751,"crush","charmbracelet\u002Fcrush","charmbracelet","Glamourous agentic coding for all 💘","",null,"Go",25220,1804,107,318,0,40,221,1058,172,119.77,"Other",false,"main",true,[27,28,29,30],"agentic-ai","ai","llms","ravishing","2026-06-12 04:00:22","# Crush\n\n\u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fstuff.charm.sh\u002Fcrush\u002Fcharm-crush.png\">\u003Cimg width=\"450\" alt=\"Charm Crush Logo\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fcf8ca3ce-8b02-43f0-9d0f-5a331488da4b\" \u002F>\u003C\u002Fa>\u003Cbr \u002F>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Freleases\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Frelease\u002Fcharmbracelet\u002Fcrush\" alt=\"Latest Release\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Factions\">\u003Cimg src=\"https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Factions\u002Fworkflows\u002Fbuild.yml\u002Fbadge.svg\" alt=\"Build Status\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">Your new coding bestie, now available in your favourite terminal.\u003Cbr \u002F>Your tools, your code, and your workflows, wired into your LLM of choice.\u003C\u002Fp>\n\u003Cp align=\"center\">终端里的编程新搭档，\u003Cbr \u002F>无缝接入你的工具、代码与工作流，全面兼容主流 LLM 模型。\u003C\u002Fp>\n\n\u003Cp align=\"center\">\u003Cimg width=\"800\" alt=\"Crush Demo\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F58280caf-851b-470a-b6f7-d5c4ea8a1968\" \u002F>\u003C\u002Fp>\n\n## Features\n\n- **Multi-Model:** choose from a wide range of LLMs or add your own via OpenAI- or Anthropic-compatible APIs\n- **Flexible:** switch LLMs mid-session while preserving context\n- **Session-Based:** maintain multiple work sessions and contexts per project\n- **LSP-Enhanced:** Crush uses LSPs for additional context, just like you do\n- **Extensible:** add capabilities via MCPs (`http`, `stdio`, and `sse`)\n- **Works Everywhere:** first-class support in every terminal on macOS, Linux, Windows (PowerShell and WSL), Android, FreeBSD, OpenBSD, and NetBSD\n- **Industrial Grade:** built on the Charm ecosystem, powering 25k+ applications, from leading open source projects to business-critical infrastructure\n\n## Installation\n\nUse a package manager:\n\n```bash\n# Homebrew\nbrew install charmbracelet\u002Ftap\u002Fcrush\n\n# NPM\nnpm install -g @charmland\u002Fcrush\n\n# Arch Linux (btw)\nyay -S crush-bin\n\n# Nix\nnix run github:numtide\u002Fnix-ai-tools#crush\n\n# FreeBSD\npkg install crush\n```\n\nWindows users:\n\n```bash\n# Winget\nwinget install charmbracelet.crush\n\n# Scoop\nscoop bucket add charm https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fscoop-bucket.git\nscoop install crush\n```\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>Nix (NUR)\u003C\u002Fstrong>\u003C\u002Fsummary>\n\nCrush is available via the official Charm [NUR](https:\u002F\u002Fgithub.com\u002Fnix-community\u002FNUR) in `nur.repos.charmbracelet.crush`, which is the most up-to-date way to get Crush in Nix.\n\nYou can also try out Crush via the NUR with `nix-shell`:\n\n```bash\n# Add the NUR channel.\nnix-channel --add https:\u002F\u002Fgithub.com\u002Fnix-community\u002FNUR\u002Farchive\u002Fmain.tar.gz nur\nnix-channel --update\n\n# Get Crush in a Nix shell.\nnix-shell -p '(import \u003Cnur> { pkgs = import \u003Cnixpkgs> {}; }).repos.charmbracelet.crush'\n```\n\n### NixOS & Home Manager Module Usage via NUR\n\nCrush provides NixOS and Home Manager modules via NUR.\nYou can use these modules directly in your flake by importing them from NUR. Since it auto detects whether its a home manager or nixos context you can use the import the exact same way :)\n\n```nix\n{\n  inputs = {\n    nixpkgs.url = \"github:NixOS\u002Fnixpkgs\u002Fnixos-unstable\";\n    nur.url = \"github:nix-community\u002FNUR\";\n  };\n\n  outputs = { self, nixpkgs, nur, ... }: {\n    nixosConfigurations.your-hostname = nixpkgs.lib.nixosSystem {\n      system = \"x86_64-linux\";\n      modules = [\n        nur.modules.nixos.default\n        nur.repos.charmbracelet.modules.crush\n        {\n          programs.crush = {\n            enable = true;\n            settings = {\n              providers = {\n                openai = {\n                  id = \"openai\";\n                  name = \"OpenAI\";\n                  base_url = \"https:\u002F\u002Fapi.openai.com\u002Fv1\";\n                  type = \"openai\";\n                  api_key = \"sk-fake123456789abcdef...\";\n                  models = [\n                    {\n                      id = \"gpt-4\";\n                      name = \"GPT-4\";\n                    }\n                  ];\n                };\n              };\n              lsp = {\n                go = { command = \"gopls\"; enabled = true; };\n                nix = { command = \"nil\"; enabled = true; };\n              };\n              options = {\n                context_paths = [ \"\u002Fetc\u002Fnixos\u002Fconfiguration.nix\" ];\n                tui = { compact_mode = true; };\n                debug = false;\n              };\n            };\n          };\n        }\n      ];\n    };\n  };\n}\n```\n\n\u003C\u002Fdetails>\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>Debian\u002FUbuntu\u003C\u002Fstrong>\u003C\u002Fsummary>\n\n```bash\nsudo mkdir -p \u002Fetc\u002Fapt\u002Fkeyrings\ncurl -fsSL https:\u002F\u002Frepo.charm.sh\u002Fapt\u002Fgpg.key | sudo gpg --dearmor -o \u002Fetc\u002Fapt\u002Fkeyrings\u002Fcharm.gpg\necho \"deb [signed-by=\u002Fetc\u002Fapt\u002Fkeyrings\u002Fcharm.gpg] https:\u002F\u002Frepo.charm.sh\u002Fapt\u002F * *\" | sudo tee \u002Fetc\u002Fapt\u002Fsources.list.d\u002Fcharm.list\nsudo apt update && sudo apt install crush\n```\n\n\u003C\u002Fdetails>\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>Fedora\u002FRHEL\u003C\u002Fstrong>\u003C\u002Fsummary>\n\n```bash\necho '[charm]\nname=Charm\nbaseurl=https:\u002F\u002Frepo.charm.sh\u002Fyum\u002F\nenabled=1\ngpgcheck=1\ngpgkey=https:\u002F\u002Frepo.charm.sh\u002Fyum\u002Fgpg.key' | sudo tee \u002Fetc\u002Fyum.repos.d\u002Fcharm.repo\nsudo yum install crush\n```\n\n\u003C\u002Fdetails>\n\nOr, download it:\n\n- [Packages][releases] are available in Debian and RPM formats\n- [Binaries][releases] are available for Linux, macOS, Windows, FreeBSD, OpenBSD, and NetBSD\n\n[releases]: https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Freleases\n\nOr just install it with Go:\n\n```\ngo install github.com\u002Fcharmbracelet\u002Fcrush@latest\n```\n\n> [!WARNING]\n> Productivity may increase when using Crush and you may find yourself nerd\n> sniped when first using the application. If the symptoms persist, join the\n> [Slack][slack] or [Discord][discord] and nerd snipe the rest of us.\n\n## Getting Started\n\nThe quickest way to get started is to grab an API key for your preferred\nprovider such as Anthropic, OpenAI, Groq, OpenRouter, or Vercel AI Gateway and just start\nCrush. You'll be prompted to enter your API key.\n\nThat said, you can also set environment variables for preferred providers.\n\n| Environment Variable        | Provider                                           |\n| --------------------------- | -------------------------------------------------- |\n| `HYPER_API_KEY`             | Charm Hyper                                        |\n| `ANTHROPIC_API_KEY`         | Anthropic                                          |\n| `OPENAI_API_KEY`            | OpenAI                                             |\n| `VERCEL_API_KEY`            | Vercel AI Gateway                                  |\n| `GEMINI_API_KEY`            | Google Gemini                                      |\n| `SYNTHETIC_API_KEY`         | Synthetic                                          |\n| `ZAI_API_KEY`               | Z.ai                                               |\n| `MINIMAX_API_KEY`           | MiniMax                                            |\n| `HF_TOKEN`                  | Hugging Face Inference                             |\n| `CEREBRAS_API_KEY`          | Cerebras                                           |\n| `OPENROUTER_API_KEY`        | OpenRouter                                         |\n| `IONET_API_KEY`             | io.net                                             |\n| `GROQ_API_KEY`              | Groq                                               |\n| `AVIAN_API_KEY`             | Avian                                              |\n| `OPENCODE_API_KEY`          | OpenCode Zen & Go                                  |\n| `VERTEXAI_PROJECT`          | Google Cloud VertexAI (Gemini)                     |\n| `VERTEXAI_LOCATION`         | Google Cloud VertexAI (Gemini)                     |\n| `AWS_ACCESS_KEY_ID`         | Amazon Bedrock (Claude)                            |\n| `AWS_SECRET_ACCESS_KEY`     | Amazon Bedrock (Claude)                            |\n| `AWS_REGION`                | Amazon Bedrock (Claude)                            |\n| `AWS_PROFILE`               | Amazon Bedrock (Custom Profile)                    |\n| `AWS_BEARER_TOKEN_BEDROCK`  | Amazon Bedrock                                     |\n| `AZURE_OPENAI_API_ENDPOINT` | Azure OpenAI models                                |\n| `AZURE_OPENAI_API_KEY`      | Azure OpenAI models (optional when using Entra ID) |\n| `AZURE_OPENAI_API_VERSION`  | Azure OpenAI models                                |\n\n### Subscriptions\n\nIf you prefer subscription-based usage, here are some plans that work well in\nCrush:\n\n- [Synthetic](https:\u002F\u002Fsynthetic.new\u002Fpricing)\n- [GLM Coding Plan](https:\u002F\u002Fz.ai\u002Fsubscribe)\n- [Kimi Code](https:\u002F\u002Fwww.kimi.com\u002Fmembership\u002Fpricing)\n- [MiniMax Coding Plan](https:\u002F\u002Fplatform.minimax.io\u002Fsubscribe\u002Fcoding-plan)\n\n### By the Way\n\nIs there a provider you’d like to see in Crush? Is there an existing model that needs an update?\n\nCrush’s default model listing is managed in [Catwalk](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcatwalk), a community-supported, open source repository of Crush-compatible models, and you’re welcome to contribute.\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcatwalk\">\u003Cimg width=\"174\" height=\"174\" alt=\"Catwalk Badge\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F95b49515-fe82-4409-b10d-5beb0873787d\" \u002F>\u003C\u002Fa>\n\n## Configuration\n\n> [!TIP]\n> Crush ships with a builtin `crush-config` skill for configuring itself. In\n> many cases you can simply ask Crush to configure itself.\n\nCrush runs great with no configuration. That said, if you do need or want to\ncustomize Crush, configuration can be added either local to the project itself,\nor globally, with the following priority:\n\n1. `.crush.json`\n2. `crush.json`\n3. `$HOME\u002F.config\u002Fcrush\u002Fcrush.json`\n\nConfiguration itself is stored as a JSON object:\n\n```json\n{\n  \"this-setting\": { \"this\": \"that\" },\n  \"that-setting\": [\"ceci\", \"cela\"]\n}\n```\n\nAs an additional note, Crush also stores ephemeral data, such as application\nstate, in one additional location:\n\n```bash\n# Unix\n$HOME\u002F.local\u002Fshare\u002Fcrush\u002Fcrush.json\n\n# Windows\n%LOCALAPPDATA%\\crush\\crush.json\n```\n\n> [!TIP]\n> You can override the user and data config locations by setting:\n>\n> - `CRUSH_GLOBAL_CONFIG`\n> - `CRUSH_GLOBAL_DATA`\n\n### LSPs\n\nCrush can use LSPs for additional context to help inform its decisions, just\nlike you would. LSPs can be added manually like so:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"lsp\": {\n    \"go\": {\n      \"command\": \"gopls\",\n      \"env\": {\n        \"GOTOOLCHAIN\": \"go1.24.5\"\n      }\n    },\n    \"typescript\": {\n      \"command\": \"typescript-language-server\",\n      \"args\": [\"--stdio\"]\n    },\n    \"nix\": {\n      \"command\": \"nil\"\n    }\n  }\n}\n```\n\n### MCPs\n\nCrush also supports Model Context Protocol (MCP) servers through three\ntransport types: `stdio` for command-line servers, `http` for HTTP endpoints,\nand `sse` for Server-Sent Events. Environment variable expansion is supported\nusing `$(echo $VAR)` syntax.\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"mcp\": {\n    \"filesystem\": {\n      \"type\": \"stdio\",\n      \"command\": \"node\",\n      \"args\": [\"\u002Fpath\u002Fto\u002Fmcp-server.js\"],\n      \"timeout\": 120,\n      \"disabled\": false,\n      \"disabled_tools\": [\"some-tool-name\"],\n      \"env\": {\n        \"NODE_ENV\": \"production\"\n      }\n    },\n    \"github\": {\n      \"type\": \"http\",\n      \"url\": \"https:\u002F\u002Fapi.githubcopilot.com\u002Fmcp\u002F\",\n      \"timeout\": 120,\n      \"disabled\": false,\n      \"disabled_tools\": [\"create_issue\", \"create_pull_request\"],\n      \"headers\": {\n        \"Authorization\": \"Bearer $GH_PAT\"\n      }\n    },\n    \"streaming-service\": {\n      \"type\": \"sse\",\n      \"url\": \"https:\u002F\u002Fexample.com\u002Fmcp\u002Fsse\",\n      \"timeout\": 120,\n      \"disabled\": false,\n      \"headers\": {\n        \"API-Key\": \"$(echo $API_KEY)\"\n      }\n    }\n  }\n}\n```\n\n### Hooks\n\nCrush has preliminary support for hooks. For details, see\n[the hook guide](.\u002Fdocs\u002Fhooks\u002F).\n\n### Ignoring Files\n\nCrush respects `.gitignore` files by default, but you can also create a\n`.crushignore` file to specify additional files and directories that Crush\nshould ignore. This is useful for excluding files that you want in version\ncontrol but don't want Crush to consider when providing context.\n\nThe `.crushignore` file uses the same syntax as `.gitignore` and can be placed\nin the root of your project or in subdirectories.\n\n### Allowing Tools\n\nBy default, Crush will ask you for permission before running tool calls. If\nyou'd like, you can allow tools to be executed without prompting you for\npermissions. Use this with care.\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"permissions\": {\n    \"allowed_tools\": [\n      \"view\",\n      \"ls\",\n      \"grep\",\n      \"edit\",\n      \"mcp_context7_get-library-doc\"\n    ]\n  }\n}\n```\n\nYou can also skip all permission prompts entirely by running Crush with the\n`--yolo` flag. Be very, very careful with this feature.\n\n### Disabling Built-In Tools\n\nIf you'd like to prevent Crush from using certain built-in tools entirely, you\ncan disable them via the `options.disabled_tools` list. Disabled tools are\ncompletely hidden from the agent.\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"disabled_tools\": [\"bash\", \"sourcegraph\"]\n  }\n}\n```\n\nTo disable tools from MCP servers, see the [MCP config section](#mcps).\n\n### Disabling Skills\n\nIf you'd like to prevent Crush from using certain skills entirely, you can\ndisable them via the `options.disabled_skills` list. Disabled skills are hidden\nfrom the agent, including builtin skills and skills discovered from disk.\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"disabled_skills\": [\"crush-config\"]\n  }\n}\n```\n\n### Agent Skills\n\nCrush supports the [Agent Skills](https:\u002F\u002Fagentskills.io) open standard for\nextending agent capabilities with reusable skill packages. Skills are folders\ncontaining a `SKILL.md` file with instructions that Crush can discover and\nactivate on demand.\n\nThe global paths we looks for skills are:\n\n* `$CRUSH_SKILLS_DIR`\n* `$XDG_CONFIG_HOME\u002Fagents\u002Fskills` or `~\u002F.config\u002Fagents\u002Fskills\u002F`\n* `$XDG_CONFIG_HOME\u002Fcrush\u002Fskills` or `~\u002F.config\u002Fcrush\u002Fskills\u002F`\n* On Windows, we _also_ look at\n  * `%LOCALAPPDATA%\\agents\\skills\\` or `%USERPROFILE%\\AppData\\Local\\agents\\skills\\`\n  * `%LOCALAPPDATA%\\crush\\skills\\` or `%USERPROFILE%\\AppData\\Local\\crush\\skills\\`\n* Additional paths configured via `options.skills_paths`\n\nOn top of that, we _also_ load skills in your project from the following\nrelative paths:\n\n* `.agents\u002Fskills`\n* `.crush\u002Fskills`\n* `.claude\u002Fskills`\n* `.cursor\u002Fskills`\n\n```jsonc\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"skills_paths\": [\n      \"~\u002F.config\u002Fcrush\u002Fskills\", \u002F\u002F Windows: \"%LOCALAPPDATA%\\\\crush\\\\skills\",\n      \".\u002Fproject-skills\",\n    ],\n  },\n}\n```\n\nYou can get started with example skills from [anthropics\u002Fskills](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills):\n\n```bash\n# Unix\nmkdir -p ~\u002F.config\u002Fcrush\u002Fskills\ncd ~\u002F.config\u002Fcrush\u002Fskills\ngit clone https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills.git _temp\nmv _temp\u002Fskills\u002F* . && rm -rf _temp\n```\n\n```powershell\n# Windows (PowerShell)\nmkdir -Force \"$env:LOCALAPPDATA\\crush\\skills\"\ncd \"$env:LOCALAPPDATA\\crush\\skills\"\ngit clone https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills.git _temp\nmv _temp\u002Fskills\u002F* . ; rm -r -force _temp\n```\n\n### Desktop notifications\n\nCrush sends desktop notifications when a tool call requires permission and when\nthe agent finishes its turn. They're only sent when the terminal window isn't\nfocused _and_ your terminal supports reporting the focus state.\n\n```jsonc\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"disable_notifications\": false, \u002F\u002F default\n  },\n}\n```\n\nTo disable desktop notifications, set `disable_notifications` to `true` in your\nconfiguration. On macOS, notifications currently lack icons due to platform\nlimitations.\n\n### Initialization\n\nWhen you initialize a project, Crush analyzes your codebase and creates\na context file that helps it work more effectively in future sessions.\nBy default, this file is named `AGENTS.md`, but you can customize the\nname and location with the `initialize_as` option:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"initialize_as\": \"AGENTS.md\"\n  }\n}\n```\n\nThis is useful if you prefer a different naming convention or want to\nplace the file in a specific directory (e.g., `CRUSH.md` or\n`docs\u002FLLMs.md`). Crush will fill the file with project-specific context\nlike build commands, code patterns, and conventions it discovered during\ninitialization.\n\n### Attribution Settings\n\nBy default, Crush adds attribution information to Git commits and pull requests\nit creates. You can customize this behavior with the `attribution` option:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"attribution\": {\n      \"trailer_style\": \"co-authored-by\",\n      \"generated_with\": true\n    }\n  }\n}\n```\n\n- `trailer_style`: Controls the attribution trailer added to commit messages\n  (default: `assisted-by`)\n  - `assisted-by`: Adds `Assisted-by: [Model Name] via Crush \u003Ccrush@charm.land>`\n    (includes the model name)\n  - `co-authored-by`: Adds `Co-Authored-By: Crush \u003Ccrush@charm.land>`\n  - `none`: No attribution trailer\n- `generated_with`: When true (default), adds `💘 Generated with Crush` line to\n  commit messages and PR descriptions\n\n### Custom Providers\n\nCrush supports custom provider configurations for both OpenAI-compatible and\nAnthropic-compatible APIs.\n\n> [!NOTE]\n> Note that we support two \"types\" for OpenAI. Make sure to choose the right one\n> to ensure the best experience!\n>\n> - `openai` should be used when proxying or routing requests through OpenAI.\n> - `openai-compat` should be used when using non-OpenAI providers that have OpenAI-compatible APIs.\n\n#### OpenAI-Compatible APIs\n\nHere’s an example configuration for Deepseek, which uses an OpenAI-compatible\nAPI. Don't forget to set `DEEPSEEK_API_KEY` in your environment.\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"providers\": {\n    \"deepseek\": {\n      \"type\": \"openai-compat\",\n      \"base_url\": \"https:\u002F\u002Fapi.deepseek.com\u002Fv1\",\n      \"api_key\": \"$DEEPSEEK_API_KEY\",\n      \"models\": [\n        {\n          \"id\": \"deepseek-chat\",\n          \"name\": \"Deepseek V3\",\n          \"cost_per_1m_in\": 0.27,\n          \"cost_per_1m_out\": 1.1,\n          \"cost_per_1m_in_cached\": 0.07,\n          \"cost_per_1m_out_cached\": 1.1,\n          \"context_window\": 64000,\n          \"default_max_tokens\": 5000\n        }\n      ]\n    }\n  }\n}\n```\n\n#### Anthropic-Compatible APIs\n\nCustom Anthropic-compatible providers follow this format:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"providers\": {\n    \"custom-anthropic\": {\n      \"type\": \"anthropic\",\n      \"base_url\": \"https:\u002F\u002Fapi.anthropic.com\u002Fv1\",\n      \"api_key\": \"$ANTHROPIC_API_KEY\",\n      \"extra_headers\": {\n        \"anthropic-version\": \"2023-06-01\"\n      },\n      \"models\": [\n        {\n          \"id\": \"claude-sonnet-4-20250514\",\n          \"name\": \"Claude Sonnet 4\",\n          \"cost_per_1m_in\": 3,\n          \"cost_per_1m_out\": 15,\n          \"cost_per_1m_in_cached\": 3.75,\n          \"cost_per_1m_out_cached\": 0.3,\n          \"context_window\": 200000,\n          \"default_max_tokens\": 50000,\n          \"can_reason\": true,\n          \"supports_attachments\": true\n        }\n      ]\n    }\n  }\n}\n```\n\n### Amazon Bedrock\n\nCrush currently supports running Anthropic models through Bedrock, with caching disabled.\n\n- A Bedrock provider will appear once you have AWS configured, i.e. `aws configure`\n- Crush also expects the `AWS_REGION` or `AWS_DEFAULT_REGION` to be set\n- To use a specific AWS profile set `AWS_PROFILE` in your environment, i.e. `AWS_PROFILE=myprofile crush`\n- Alternatively to `aws configure`, you can also just set `AWS_BEARER_TOKEN_BEDROCK`\n\n### Vertex AI Platform\n\nVertex AI will appear in the list of available providers when `VERTEXAI_PROJECT` and `VERTEXAI_LOCATION` are set. You will also need to be authenticated:\n\n```bash\ngcloud auth application-default login\n```\n\nTo add specific models to the configuration, configure as such:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"providers\": {\n    \"vertexai\": {\n      \"models\": [\n        {\n          \"id\": \"claude-sonnet-4@20250514\",\n          \"name\": \"VertexAI Sonnet 4\",\n          \"cost_per_1m_in\": 3,\n          \"cost_per_1m_out\": 15,\n          \"cost_per_1m_in_cached\": 3.75,\n          \"cost_per_1m_out_cached\": 0.3,\n          \"context_window\": 200000,\n          \"default_max_tokens\": 50000,\n          \"can_reason\": true,\n          \"supports_attachments\": true\n        }\n      ]\n    }\n  }\n}\n```\n\n### Local Models\n\nLocal models can also be configured via OpenAI-compatible API. Here are two common examples:\n\n#### Ollama\n\n```json\n{\n  \"providers\": {\n    \"ollama\": {\n      \"name\": \"Ollama\",\n      \"base_url\": \"http:\u002F\u002Flocalhost:11434\u002Fv1\u002F\",\n      \"type\": \"openai-compat\",\n      \"models\": [\n        {\n          \"name\": \"Qwen 3 30B\",\n          \"id\": \"qwen3:30b\",\n          \"context_window\": 256000,\n          \"default_max_tokens\": 20000\n        }\n      ]\n    }\n  }\n}\n```\n\n#### LM Studio\n\n```json\n{\n  \"providers\": {\n    \"lmstudio\": {\n      \"name\": \"LM Studio\",\n      \"base_url\": \"http:\u002F\u002Flocalhost:1234\u002Fv1\u002F\",\n      \"type\": \"openai-compat\",\n      \"models\": [\n        {\n          \"name\": \"Qwen 3 30B\",\n          \"id\": \"qwen\u002Fqwen3-30b-a3b-2507\",\n          \"context_window\": 256000,\n          \"default_max_tokens\": 20000\n        }\n      ]\n    }\n  }\n}\n```\n\n## Logging\n\nSometimes you need to look at logs. Luckily, Crush logs all sorts of\nstuff. Logs are stored in `.\u002F.crush\u002Flogs\u002Fcrush.log` relative to the project.\n\nThe CLI also contains some helper commands to make perusing recent logs easier:\n\n```bash\n# Print the last 1000 lines\ncrush logs\n\n# Print the last 500 lines\ncrush logs --tail 500\n\n# Follow logs in real time\ncrush logs --follow\n```\n\nWant more logging? Run `crush` with the `--debug` flag, or enable it in the\nconfig:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"debug\": true,\n    \"debug_lsp\": true\n  }\n}\n```\n\n## Provider Auto-Updates\n\nBy default, Crush automatically checks for the latest and greatest list of\nproviders and models from [Catwalk](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcatwalk),\nthe open source Crush provider database. This means that when new providers and\nmodels are available, or when model metadata changes, Crush automatically\nupdates your local configuration.\n\n### Disabling automatic provider updates\n\nFor those with restricted internet access, or those who prefer to work in\nair-gapped environments, this might not be want you want, and this feature can\nbe disabled.\n\nTo disable automatic provider updates, set `disable_provider_auto_update` into\nyour `crush.json` config:\n\n```json\n{\n  \"$schema\": \"https:\u002F\u002Fcharm.land\u002Fcrush.json\",\n  \"options\": {\n    \"disable_provider_auto_update\": true\n  }\n}\n```\n\nOr set the `CRUSH_DISABLE_PROVIDER_AUTO_UPDATE` environment variable:\n\n```bash\nexport CRUSH_DISABLE_PROVIDER_AUTO_UPDATE=1\n```\n\n### Manually updating providers\n\nManually updating providers is possible with the `crush update-providers`\ncommand:\n\n```bash\n# Update providers remotely from Catwalk.\ncrush update-providers\n\n# Update providers from a custom Catwalk base URL.\ncrush update-providers https:\u002F\u002Fexample.com\u002F\n\n# Update providers from a local file.\ncrush update-providers \u002Fpath\u002Fto\u002Flocal-providers.json\n\n# Reset providers to the embedded version, embedded at crush at build time.\ncrush update-providers embedded\n\n# For more info:\ncrush update-providers --help\n```\n\n## Metrics\n\nCrush records pseudonymous usage metrics (tied to a device-specific hash),\nwhich maintainers rely on to inform development and support priorities. The\nmetrics include solely usage metadata; prompts and responses are NEVER\ncollected.\n\nDetails on exactly what’s collected are in the source code ([here](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Ftree\u002Fmain\u002Finternal\u002Fevent)\nand [here](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Fblob\u002Fmain\u002Finternal\u002Fllm\u002Fagent\u002Fevent.go)).\n\nYou can opt out of metrics collection at any time by setting the environment\nvariable by setting the following in your environment:\n\n```bash\nexport CRUSH_DISABLE_METRICS=1\n```\n\nOr by setting the following in your config:\n\n```json\n{\n  \"options\": {\n    \"disable_metrics\": true\n  }\n}\n```\n\nCrush also respects the [`DO_NOT_TRACK`](https:\u002F\u002Fdonottrack.sh\u002F) convention\nwhich can be enabled via `export DO_NOT_TRACK=1`.\n\n## Q&A\n\n### Why is clipboard copy and paste not working?\n\nInstalling an extra tool might be needed on Unix-like environments.\n\n| Environment         | Tool                     |\n| ------------------- | ------------------------ |\n| Windows             | Native support           |\n| macOS               | Native support           |\n| Linux\u002FBSD + Wayland | `wl-copy` and `wl-paste` |\n| Linux\u002FBSD + X11     | `xclip` or `xsel`        |\n\n## Contributing\n\nSee the [contributing guide](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush?tab=contributing-ov-file#contributing).\n\n## Whatcha think?\n\nWe’d love to hear your thoughts on this project. Need help? We gotchu. You can find us on:\n\n- [Twitter](https:\u002F\u002Ftwitter.com\u002Fcharmcli)\n- [Slack][slack]\n- [Discord][discord]\n- [The Fediverse](https:\u002F\u002Fmastodon.social\u002F@charmcli)\n- [Bluesky](https:\u002F\u002Fbsky.app\u002Fprofile\u002Fcharm.land)\n\n[slack]: https:\u002F\u002Fcharm.land\u002Fslack\n[discord]: https:\u002F\u002Fcharm.land\u002Fdiscord\n\n## License\n\n[FSL-1.1-MIT](https:\u002F\u002Fgithub.com\u002Fcharmbracelet\u002Fcrush\u002Fraw\u002Fmain\u002FLICENSE.md)\n\n---\n\nPart of [Charm](https:\u002F\u002Fcharm.land).\n\n\u003Ca href=\"https:\u002F\u002Fcharm.land\u002F\">\u003Cimg alt=\"The Charm logo\" width=\"400\" src=\"https:\u002F\u002Fstuff.charm.sh\u002Fcharm-banner-softy.jpg\" \u002F>\u003C\u002Fa>\n\n\u003C!--prettier-ignore-->\nCharm热爱开源 • Charm loves open source\n","Crush 是一个旨在通过终端为开发者提供编程辅助的工具。它支持多种大型语言模型（LLMs），允许用户在会话中灵活切换模型并保持上下文连贯，同时基于LSP协议增强代码理解和处理能力。此外，Crush具有高度可扩展性，可以通过MCPs添加更多功能，并且跨平台兼容性强，支持macOS、Linux、Windows等操作系统。此项目特别适合需要高效编码环境、希望集成AI助手以提高开发效率的程序员使用。",2,"2026-06-11 03:00:22","top_language"]