6bc360d375
Allow yaml ( #60 )
2025-06-09 16:24:27 +03:00
e37c7952dd
create agent from scratch for better control ( #59 )
2025-06-09 15:46:17 +03:00
8645a8c36a
Huge refactor ( #58 )
...
* overhaul
* fix ui glitch
* system messages
* use sonnet 4
2025-06-09 14:38:31 +03:00
6c8f1114e8
update .gitignore
2025-06-09 12:11:17 +03:00
7a81dc3ccf
upgrade mcp-go
2025-06-08 13:45:47 +03:00
b4750c5852
Document SSE ( #44 )
2025-05-13 23:22:59 +03:00
009beb5b64
Quick hack to enable image support returned from tools ( #45 )
...
Co-authored-by: spsobole <spsobole@thirdmartini.com >
2025-05-13 23:22:30 +03:00
494851a71f
fix: crash when sending tool_result with anthropic and google ( #34 )
...
* fix: tool result handling in anthropic
* fix: set the role outside the loop
* fix: role mapping in implementation of google provider and anthropic provider
* fix: simplify role mapping logic and update role constants
2025-04-25 15:13:55 +03:00
beaa623fc5
feat: support system prompt ( #29 )
...
* support system prompt for ollama
* support system prompt for openai
* support system prompt for google gemini
* fixed setting for google gemini system-prompt
* support system prompt for anthropic
* edit README
* edit commments
* fix system-prompt flag desc
* modifed var name for systemPromp
2025-04-23 17:39:21 +03:00
4facfd6151
fix: Do not panic when reading malformed properties ( #31 )
...
* fix: Do not panic when reading malformed properties
Signed-off-by: Glenn Lewis <6598971+gmlewis@users.noreply.github.com >
* Add one more check
Signed-off-by: Glenn Lewis <6598971+gmlewis@users.noreply.github.com >
---------
Signed-off-by: Glenn Lewis <6598971+gmlewis@users.noreply.github.com >
2025-04-23 17:38:01 +03:00
c47f090061
Add number and integer types to google type converter ( #32 )
2025-04-23 17:37:23 +03:00
4fe66b099e
Fix multiple requests at the same time ( #33 )
...
Previously when multiple tool calls were requested by the LLM we got the
error `An assistant message with 'tool_calls' must be followed by tool
messages responding to each 'tool_call_id'.` although the calls were
processed.
Apparently each response needs to be its own message, rather than its
own block.
Closes #7
2025-04-23 17:37:01 +03:00
ec996fd1e1
Add support of config for SSE servers. Add Authorization header support ( #24 )
...
* Add support of config for SSE servers. Add Authorization header support
* Change the format of JSON for SSE servers to be similar to other tools
---------
Co-authored-by: Roman Gelembjuk <Roman Gelembjuk>
2025-04-16 20:18:44 +03:00
c572625257
Add ``omitempty` to `reasoning_content`` in OpenAI message API ( #27 )
...
Add ```omitempty``` to ```reasoning_content``` in OpenAI message API. See issue [link](https://github.com/mark3labs/mcphost/issues/23 )
2025-04-16 20:18:12 +03:00
f0be8458aa
Increase maximum length of prompt input ( #25 )
...
Default maximum length is 400 characters.
2025-04-16 20:17:39 +03:00
3076686b18
Add a Discord link\
2025-04-12 12:55:53 +03:00
b4057a917a
Allow to connect to remote SSE MCP server ( #22 )
...
Co-authored-by: Roman Gelembjuk <Roman Gelembjuk>
2025-04-11 23:22:53 +03:00
f1ed1fcb5c
Add explanation about Ollama environment variables ( #17 )
...
Co-authored-by: Ed Zynda <ezynda3@users.noreply.github.com >
2025-04-11 23:21:16 +03:00
d8583a7883
Fix default config file name in help message ( #16 )
...
* Fix default config file name
* Update README.md
2025-04-11 23:19:33 +03:00
d58251f53e
Add Google/Gemini support ( #15 )
...
* add google/gemini support
* fix user/model role issues, use chat session
* properly maintain history
* update gemini/google provider, pipe context through runPrompt, minor fixes
* minor formatting fix in usage tooltip
* update readme to showcase Google/Gemini support
---------
Co-authored-by: Ed Zynda <ezynda3@users.noreply.github.com >
2025-04-11 23:19:06 +03:00
54ff95c421
add contribute guide and fix openai implement to openai-compatible ( #13 )
2025-04-11 23:16:43 +03:00
e614a4d061
Merge branch 'main' of github.com:mark3labs/mcphost
2025-02-23 12:16:30 +03:00
32a4831179
use background context
2025-02-23 12:16:17 +03:00
9681971ee9
Update README
2025-01-03 07:12:20 +03:00
c4ed189f48
Setup goreleaser
2025-01-03 07:08:49 +03:00
88a8c7bf15
support anthropicBaseURL , openaiAPIKey, anthropicAPIKey cli params ( #6 )
...
* * support anthropicBaseURL , openaiAPIKey, anthropicAPIKey cli params
* * support anthropicBaseURL , openaiAPIKey, anthropicAPIKey cli params
* opt
* remove the non-English comments
* add new Flags to document
2024-12-27 14:10:29 +03:00
b32fd143c6
cleanup
2024-12-21 20:28:35 +03:00
9a50bcbcbe
update versions and formatting
2024-12-20 18:25:10 +03:00
e1a974511d
remove patch version
2024-12-20 18:21:03 +03:00
5c9246cd5f
Update README
2024-12-20 18:20:05 +03:00
5ac7908227
Refactor ( #5 )
...
Refactoring so hard!
2024-12-20 18:15:14 +03:00
e3a64a107d
update SDK version
2024-12-18 11:31:27 +03:00
fa7cbb0b55
Fix ollama
2024-12-11 17:13:41 +03:00
0bbe5f15c1
Update deps
2024-12-11 15:36:01 +03:00
e69bbb8354
Update deps
2024-12-11 15:35:56 +03:00
5201b57ec6
Update README
2024-12-10 22:29:08 +03:00
e7cebe73da
Update README and LICENSE
2024-12-10 22:25:15 +03:00
b79b521911
Add message window to prevent overloading context
2024-12-10 22:19:35 +03:00
57fe2c6795
Update styling
2024-12-10 21:52:33 +03:00
ea68e00db6
Add /history command
2024-12-10 18:08:17 +03:00
87ffdec8f0
Add usage stats and handle overloaded errors
2024-12-10 17:40:02 +03:00
2d79e5989c
Remove anthropic go sdk dependency
2024-12-10 17:32:12 +03:00
6b2180ac7c
Fix tool check
2024-12-10 16:36:13 +03:00
8b0001184f
Disable tools for models that do not support it.
2024-12-09 16:46:12 +03:00
8350b7a55a
Update MCP-Go SDK
2024-12-09 10:12:51 +03:00
986f5560c5
Refactor
2024-12-09 00:01:06 +03:00
8301aa1a68
Update README
2024-12-08 19:37:30 +03:00
aca9d750ba
update config location and add LICENSE
2024-12-08 19:33:45 +03:00
d80a14f660
Initial commit
2024-12-08 19:27:53 +03:00
Ed Zynda