Skip to content

Conversation

@github-actions
Copy link
Contributor

🤖 I have created a release beep boop

12.0.0 (2025-02-27)

⚠ BREAKING CHANGES

  • agents: agents are now in strategies.chat.agents
  • chat: ✨ can pin buffers and files to requests

Features

  • ✨ add support for snacks.nvim (#741) (5407aba)
  • ✨ Agentic Workflows and much more (#832) (f26fa4e)
  • ✨ can edit setting and message history (#753) (2592d26)
  • ✨ workspace slash command (#702) (8173b5d)
  • #604 send_code can be a function (30b899e)
  • adapter and debug window improvement (be55c9b)
  • adapters: add claude 3.7 thought to copilot adapter (#1001) (6e53570)
  • adapters: add claude-3.7-sonnet and make default (#996) (c207203)
  • adapters: add claude-3.7-sonnet to copilot (#1002) (9729441)
  • adapters: add DeepSeek Adapter (#683) (6668b5c)
  • adapters: add extended thinking and extended output for claude-3.7 (#998) (a497adb)
  • adapters: add gemini-2.0-pro-exp-02-05 model option (#1007) (bef6d90)
  • adapters: add GitHub Models support (#862) (4a6567b)
  • adapters: add Hugging Face adapter (#527) (48747c4)
  • adapters: add reasoning effort to copilot and openai (#812) (b9a58eb)
  • adapters: can hide default adapters (#925) (f4b72d7)
  • adapters: dynamically get available models from copilot endpoint (#1012) (d62c8e1)
  • adapters: update copilot and openai adapters for o1 models (c7ba92a)
  • add new events (9b85331)
  • add reference handling to workflow processing (#969) (4c9f120)
  • add symbols for scala (#768) (4535cb5)
  • chat:blink.cmp support (cb2c93f)
  • chat: ✨ can pin buffers and files to requests (fc8ee3a)
  • chat: ✨ can watch buffers (176c13e)
  • chat: ✨ LLM header name can now be a function (#778) (5a82a54)
  • chat: add CodeCompanionChatOpened event (fdfc640)
  • chat: adding to chat initializes buffer if needed (#567) (8c82ced)
  • chat: can delete references (367eccc)
  • chat: can delete references (c0b8c25)
  • chat: customize split position (#556) (7374192)
  • chat: easier to add refs from external sources (#960) (b7b43f6)
  • chat: keymaps can be custom functions (5f729f9)
  • chat: ui title is configurable (8c93ab0)
  • completion: auto setup blink.cmp (90a6603)
  • completion: suggestions in blink have better score (4317513)
  • copilot: add o3-mini support (#810) (2816d12)
  • copilot: add support for gemini 2.0 flash (#855) (#856) (c0ea1c0)
  • copilot: add token count (4fd26a6)
  • deepseek: #692 #669 show reasoning output (#716) (b4467a1)
  • event: add CodeCompanionChatPin event (f86f4a7)
  • events: add strategy and adapter to payload (#822) (b2e06de)
  • events: new request streaming event (#878) (6846482)
  • inline: new and improved assistant (#904) (3e4fb61)
  • prompts: ✨ can now add references to the prompt library (d5d86c9)
  • prompts: system prompt is optional for custom prompts (90820fc)
  • register source for filetype in blink.cmp (#661) (2214916)
  • slash_cmd: notify if no terminal buffer is found (3687b88)
  • slash_cmds: buffer slash command now sends line numbers (25dab99)
  • slash_cmds: can be functions and modules (#844) (59ea59e)
  • slash_cmds: terminal command only shows latest changes (#898) (cde1882)
  • slash_command: control system prompts in /workspace (#736) (16d7f62)
  • slash_commands: help cmd now prompts user to trim if exceeding max_lines (33c326a)
  • tools: [@files](https://github.com/files) tool can now read specific lines (26c5b57)
  • tools: files tool now uses search/replace block to edit text (a02f6c7)
  • tools: allow passing a table as the callback of a tool. (#936) (689ea6e)
  • tools: when stderr is not empty, do not discard stdout when exit code is 0 (#955) (d8de491)
  • workspaces: ✨ now supporting URLs (#808) (5a1212d)
  • workspaces: handle multiple workflows better (#806) (e463d65)

Bug Fixes

  • #501 blink.compat causing issues with completion menu (c5065d0)
  • #542 use vim.treesitter instead of nvim-treesitter. (#543) (b741490)
  • #698 custom variables (#711) (104f913)
  • #732 can use deepseek with no config (#733) (23d81a9)
  • #769 ollama adapter (#793) (8a98c71)
  • adapter: #731 openai_compat adapter calling API too much (#738) (184ca40)
  • adapter: #769 ollama adapter model not being expanded (6a78557)
  • adapters: #528 Gemini adapter fails if system prompt is removed (7ca4364)
  • adapters: #560 nil url in openai_compatible adpater (5cc2693)
  • adapters: on_exit is called when request is cancelled (#942) (df2c5f6)
  • adapters: openai_compatible no longer forces schema (#914) (ba06239), closes #918
  • adapters: dynamically getting copilot adapters when switching adapters (c0a820e)
  • adapters: further openai_compatible fixes (a84f6ff)
  • adapters: gemini-2.0-flash-exp inline strategy (#568) (a9faa69)
  • adapters: handle multiline userdata in copilot adapter (#770) (4282c68)
  • adapters: inline editing with copilot closes #1036 #1027 (#1037) (aae4ad7)
  • adapters: max_tokens error with anthropic (#1031) (59c9109)
  • adapters: only try and replace str variables (#999) (c5990b4)
  • adapters: openai_compatible (0d1388f)
  • adapters: openai_compatible hardcoded models endpoint closes #1022 (#1038) (4585695)
  • adapter: variable substitution in 'raw'. (#980) (6c628f5)
  • allow for multiple completions in a single line (3516e46)
  • blink: #587 slash commands (1b8ab18)
  • blink: slash commands leaving brackes in chat buffer (820bc52)
  • chat: #622 parse messages when show_header_separator is on (6040ce4)
  • chat: #783 display settings in chat buffer (#784) (cd712c5)
  • chat: #814 show_header_separator (#816) (38793d3)
  • chat: #887 header separator is now detected (#902) (63e76c5)
  • chat: #889 pinned file references always have same id (#890) (cbac9bc)
  • chat: #906 ignore system prompt now works (#909) (4d0e4dd)
  • chat: #915 parsing show_settings when InsertLeave is executed (#916) (1103551)
  • chat: agent's tools are no displayed individually (d889210)
  • chat: always strip references from messages (4c91a46)
  • chat: better cancelling of requests (#828) (84e8098)
  • chat: can debug adapters with schema.model.default as a function (#797) (fbb15a8)
  • chat: change deepseek models (7e211d5)
  • chat: changing adapter in empty chat w/o system prompt (#616) (3e64e6c)
  • chat: changing adapters for models with table type (3a961bb)
  • chat: debug window height and width asfunctions (#926) (1d67908), closes #905
  • chat: incorrectly applying settings (#881) (2b0fe19)
  • chat: message editing for adapters with model schema as a function (b6dcc37)
  • chat: more graceful cancellations (#739) (05c8853)
  • chat: navigate between H2 headers in buffer only (#781) (391aeae)
  • chat: prevent double adding of references (953b20d)
  • chat: prevent duplicate references (fb6444b)
  • chat: show documentation for blink.cmp (f41dbab)
  • chat: slash commands from prompt library not visible (3376f60)
  • cmd: strategy now respects vim.g.codecompanion_adapter (b629005)
  • cmp: extend default sources (8b0ebb4)
  • completion: #597 blink configuration (a1166d5)
  • completion: blink.cmp failing for slash commands (6dd3309)
  • copilot: show token output in chat (6908fd7)
  • deepseek: #760 tooling and update adapter utils (3d3b1a1)
  • deepseek: messages order (#700) (c911a8e)
  • delete diff buffers instead of closing (#852) (4c48bd3)
  • gemini: remove cycle from system prompt (d6b95bb)
  • http: prevent model from no schema.model (#839) (abaf645)
  • inline: #516 switching adapters via vim.g.codecompanion_adapter (0d91c08)
  • inline: #531 inline edits in c++ buffers (73e8ea5)
  • inline: #613 negative column indexing (977a521)
  • inline: #688 end_col out of range (6d2a477)
  • inline: ensure placement is in lower case (#949) (04019d5)
  • inline: setting placement from prompt library closes #990 (#991) (836dafc)
  • inline: tweak system prompt (#754) (4985375)
  • logging in plugin (#786) (ec30170)
  • ollama: allow users to customize the authorization header in get_models closes #976 (#985) (a4f1046)
  • openai: #619 o1 streaming (83786d8)
  • openai: #619 o1 streaming (8bfb496)
  • openai: #892 handle different finish reasons (#893) (70179b0)
  • openai: model streaming (#896) (28a03a3)
  • slash_cmd: improve symbols notifcations (e6b36b4)
  • slash_cmds: file slash command will show hidden files with telescope (39c7ca0)
  • slash_command: create cache folder on windows for /fetch (#756) (223e35f)
  • slash_commands: #515 help tags line limit can be customised (e1b9876)
  • slash_commands: #803 mini_pick provider (#804) (4a895b9)
  • slash_commands: handle no content in terminal slash command (aae2085)
  • slash_commands: prompt library now works with blink.cmp (1027f47)
  • slash_commands: scope constants to be local (#787) (7daf143)
  • slash_command: stop double wrapping output from fetch (c53a917)
  • slash_command: typescript ft not recognised (#723) (c114e42)
  • tools: #503 shell redirection (7023e58)
  • tools: add missing write flag attribute when editing files (#966) (d5bc944)
  • tools: better error handling (0d8de44)
  • tools: can now chain jobs effectively (#945) (2bab6d7)
  • tools: explictly add cwd to plenary.job call (e78006d)
  • tools: make stderr capturing work for "streaming" output (#943) (199824e)
  • tools: output correctly in the chat buffer (3a6acfb)
  • tools: remove ANSI sequences from all output (2836849)
  • tools: strip ansi from cmd output (679a9a5)
  • use slash commands with :CodeCompanion cmd (8cbf696)
  • utils: #488 fix edge cases in visual selection (84a8e89)
  • utils: parse "boolean_scalar" fields when decoding yaml (#635) (3bf8cac)
  • workflow: can auto_submit the first prompt (86f96d5)
  • workflows: wrap in vim.schedule for performance (f0f6e26)
  • workspaces: add description to urls from cache and check they exist (#900) (2d08176)
  • workspaces: workspace path constant cleared after first use (#761) (b39acc1)
  • yaml: handle nil values in show_settings (9b44789)

Performance Improvements

  • chat: #552 improve settings tree-sitter query (#708) (c7e54d2)
  • http: improve resiliency of requests (d99928b)

Miscellaneous Chores

Code Refactoring

  • agents: agents are now in strategies.chat.agents (3f437a6)

This PR was generated with Release Please. See documentation.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

2 participants