Ling0925
/
suna
mirror of https://github.com/kortix-ai/suna.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
1,681 Commits 30 Branches 7 Tags 406 MiB
Commit Graph

64 Commits

Author SHA1 Message Date
Cursor Agent 9e0e89ff61 Add auto-continue for length and context window finish reasons 
Co-authored-by: sharath <sharath@kortix.ai>
 2025-07-19 22:35:54 +00:00
Sharath 8cbaf3d44a
fix(backend): anthropic overloaded fallback failure (#929) 2025-07-11 18:36:15 +05:30
marko-kraemer b0ab413866 fix: optimize blocking command execution with marker-based completion detection 
- Use unique completion markers instead of unreliable prompt detection
- Reduce polling from 2s to 0.5s intervals for faster response
- Fix SessionExecuteRequest import path
- Commands now return immediately upon completion vs waiting full timeout
 2025-07-06 20:07:26 +02:00
marko-kraemer dc5496ae94 wip 2025-07-06 16:40:27 +02:00
marko-kraemer 6a6b9d8e85 refactor wip 2025-07-05 18:07:35 +02:00
Soumyadas15 18057a4db6 chore(dev): agent agnostic threads 2025-06-26 16:59:04 +05:30
sharath 38f4355838
fix(thread_manager): implement batching for message retrieval to enhance database performance and mitigate db limits 2025-06-22 16:42:55 +00:00
sharath 6501543814
feat(thread_manager): implement middle-out message compression to optimize message handling 2025-06-19 18:05:45 +00:00
sharath 184b54fc8f
fix(thread_manager): enhance content validation to include interactive elements 2025-06-19 14:19:48 +00:00
sharath e8dbc55165
fix(thread_manager): add message compression by omitting messages to handle token limits 2025-06-18 21:25:32 +00:00
sharath 05a80e5f55
fix(thread_manager): adjust truncation logic and improve token compression handling 2025-06-18 07:25:29 +00:00
sharath 3507230ed0
fix(compress_messages): truncate the middle of last msg and reduce token size 2025-06-18 05:53:51 +00:00
Sharath 85cb4b032b
Update backend/agentpress/thread_manager.py 2025-06-16 14:39:13 +05:30
sharath e7ef1c64c5
fix(thread_manager): enhance message content parsing and ensure proper compression of prepared messages 2025-06-16 09:07:47 +00:00
sharath 2a49ad8936
fix(thread_manager): improve message handling and update max_tokens calculation 2025-06-16 07:06:24 +00:00
sharath 0e41358047
fix(context): context window compression 2025-06-15 21:14:09 +00:00
sharath c00497865f
fix(thread_manager): adjust max_tokens calculation for deepseek model 2025-06-06 18:48:55 +00:00
sharath b5ae395fed
refactor(thread-manager): standardize max_tokens parameter in message compression methods 2025-06-06 09:18:34 +00:00
sharath 93660e4a61
feat(thread-manager): add message compression methods to handle long messages efficiently 2025-06-06 08:56:14 +00:00
Bobbie 1b269a96b6
Merge pull request #658 from escapade-mckv/prod-2a 
Prod 2a
 2025-06-06 13:21:29 +05:30
Sharath aee5945907
hotfix(context-limit): truncate more content to allow more messages 2025-06-06 12:36:11 +05:30
sharath 11e8b4f7bd
Merge branch 'PRODUCTION' into sync/production 2025-06-05 06:48:53 +00:00
sharath eda0c69bea
fix(context-limit-retry): don't retry when context limit reached 2025-06-05 06:46:27 +00:00
Soumyadas15 8a7c33c86b chore(dev): thread compression for new format 2025-06-03 12:02:50 +05:30
Soumyadas15 b42e78b598 chore(dev): agent builder 2025-05-31 21:01:20 +05:30
sharath c14a2e4cb4
fix(thread_manager): reduce maximum message content length from 300k to 200k characters to prevent context overload 2025-05-29 15:50:29 +00:00
sharath 0bfd7ecb3f
fix(thread_manager): adjust token count condition to handle default max token value 2025-05-29 15:28:53 +00:00
sharath 16ba4f7c36
fix(thread_manager): update message selection query to include only message_id and content 2025-05-29 14:20:57 +00:00
sharath 960dd671f2
feat(context-compression): add ExpandMessageTool for message expansion and integrate into agent run process 2025-05-29 13:47:31 +00:00
Soumyadas15 fb6c130136 chore(dev): fix double escape isue 2025-05-25 02:34:07 +05:30
sharath 7db395dd88
refactor(api): enhance trace handling in response processing 2025-05-24 12:50:36 +00:00
sharath ff1670be90
feat(backend): langfuse traces 2025-05-21 00:39:28 +00:00
marko-kraemer c6d13b5187 run, thread_run, response processor, raise errs properly 2025-05-10 03:58:57 +02:00
marko-kraemer 7ff15d8dc7 make non-anthropic model work with response example, fe padding bottom increase cuz of new chat input 2025-05-07 01:56:52 +02:00
marko-kraemer 12704f2786 fix 2025-05-05 02:53:24 +02:00
Korzhavin Ivan 68fafdd385 Refactor agent and tool configurations to support new model names and parameters 
- Adjusted max_tokens logic based on model type, including new conditions for "gpt-4".
- Enhanced MessageTool to include detailed documentation and examples for user interactions.
- Improved ThreadManager to handle new tool choices and reasoning efforts.
- Updated LLM service to incorporate enable_thinking and reasoning_effort parameters.
- Modified chat input component to dynamically manage model selection and thinking options.
 2025-05-03 14:17:46 +02:00
marko-kraemer 973861b2d2 wip 2025-05-03 02:26:25 +02:00
marko-kraemer f9ca36efa0 wip 2025-04-23 09:13:07 +01:00
LE Quoc Dat 17e3e0e68c Merge branch 'main' into dat_context 2025-04-18 17:01:59 +01:00
LE Quoc Dat 06f1782f4b fix nasty bug xN system prompt 2025-04-18 09:09:09 +01:00
LE Quoc Dat 8669e40312 disable context manager 2025-04-18 07:13:44 +01:00
marko-kraemer 582f0d228b unified message structure on sse stream & message get thread 2025-04-18 06:22:06 +01:00
LE Quoc Dat c84ee59dc6 thinking & reasoning 2025-04-18 05:49:41 +01:00
LE Quoc Dat 9beb4dc215 fix non-stream async 2025-04-18 04:41:55 +01:00
marko-kraemer 647e10a74b wip 2025-04-17 23:17:48 +01:00
Adam Cohen Hillel 6a59b8e110 token manager 2025-04-17 00:17:09 +01:00
marko-kraemer 9fbad13468 wip 2025-04-16 23:13:04 +01:00
marko-kraemer 31205b0320 agent wip 2025-04-11 00:53:01 +01:00
marko-kraemer 57d08c76af wip 2025-04-10 17:02:21 +01:00
marko-kraemer cff9ba57f6 include xml examples, agent wip 2025-04-10 14:13:32 +01:00