Commit Graph
5081 Commits
Author SHA1 Message Date
Sydney Runkle b81c21f311 docs updates 2025-04-23 17:48:47 -07:00
Sydney Runkle cd6b086374 await 2025-04-23 16:36:37 -07:00
Sydney Runkle 5929b0b08d more test fixes 2025-04-23 16:19:46 -07:00
Sydney Runkle 9f33b4e529 fix test to reflect bug fix 2025-04-23 16:03:08 -07:00
Sydney RunkleandGitHub f0fb8d9187 Merge branch 'main' into sr/better-interrupts 2025-04-23 15:58:57 -07:00
Sydney RunkleandGitHub 0ff913c50b Fix double interrupt return caused by dynamic tasks (#4389)
Fix issue found in
https://github.com/langchain-ai/langgraph/pull/4374#discussion_r2056701206
2025-04-23 15:56:47 -07:00
Sydney RunkleandGitHub 67075bfb7f Update libs/langgraph/tests/test_pregel_async.py 2025-04-23 15:50:22 -07:00
Sydney Runkle 37b2956758 helpful comments + test 2025-04-23 14:24:57 -07:00
Sydney Runkle f42fc971e3 fix double interrupt raise with task path bool flag 2025-04-23 14:16:35 -07:00
Vadym BardaandGitHub 6bcab08f55 langgraph: release 0.3.33 (#4388) 2025-04-23 16:28:12 -04:00
Sydney Runkle 8fe7c8101f more explicit tests 2025-04-23 11:49:26 -07:00
Vadym BardaandGitHub 6bca615e6a langgraph: allow nested lists/dicts of primitives in RemoteGraph config (#4387) 2025-04-23 14:04:54 -04:00
Vadym BardaandGitHub 0147790937 langgraph: release 0.3.32 (#4386) 0.3.32 2025-04-23 11:51:29 -04:00
Vadym BardaandGitHub 364c572f0f langgraph: fix messages streaming for list of Commands (#4379)
Fixes #4372
2025-04-23 11:46:57 -04:00
langchain-infraandGitHub 8af4b63c69 docs: Add docs for REDIS_KEY_PREFIX, REDIS_CLUSTER env vars (#4334)
# Description 

Add documentation for using the REDIS_KEY_PREFIX, REDIS_CLUSTER env
vars.
2025-04-23 11:17:03 -04:00
Asamu DavidandGitHub 941b4b1287 Merge branch 'main' into da/4-17/add-redis-prefix-support 2025-04-23 15:52:30 +01:00
Asamu DavidandGitHub 24f04fba37 Update REDIS_CLUSTER env doc.md 2025-04-23 15:52:07 +01:00
William FHandGitHub 05f21a384b Add image arg to up command (#4385)
Using this argument, you can get more customization since you can do
`langgraph build` or directly `docker build` your image and then re-use
the `langgraph up --image my-image` and have it also spin up redis &
postgres for you.

Easier then writing your own compose file
cli==0.2.7
2025-04-23 07:48:01 -07:00
Vadym BardaandGitHub 29e9ee2d7b docs(agents): use list of messages format (#4378) 2025-04-23 01:25:00 +00:00
Sydney Runkle d6b4ee348f goodness, last test fix 2025-04-22 17:32:13 -07:00
Sydney Runkle 9e89b68596 fixing prebuilt tests 2025-04-22 17:24:54 -07:00
Sydney Runkle 1251eeaa48 final test fixes, hopefully 2025-04-22 17:21:56 -07:00
Sydney Runkle 85e73653f9 linting and test fixes 2025-04-22 17:10:54 -07:00
Sydney RunkleandGitHub a8b70cb23c Merge branch 'main' into sr/better-interrupts 2025-04-22 13:48:03 -07:00
Sydney RunkleandGitHub 173627a94c langgraph[lint]: Upgrade to Python 3.9+ syntax (#4368)
Also added "UP" (pyupgrade) rule to `pyproject.toml`
2025-04-22 13:47:18 -07:00
Sydney Runkle edd7d608cd final linting 2025-04-22 13:40:53 -07:00
Sydney Runkle 8977a35060 linting post merge 2025-04-22 13:01:40 -07:00
Sydney RunkleandGitHub 0a6e5a18bb Merge branch 'main' into pyupgrade-39 2025-04-22 12:30:53 -07:00
Sydney Runkle abc6323c6c Merge branch 'sr/better-interrupts' of https://github.com/langchain-ai/langgraph into sr/better-interrupts 2025-04-22 11:39:24 -07:00
Sydney Runkle d5a1bb05f5 revert changes to lockfile 2025-04-22 11:35:56 -07:00
Sydney Runkle 8506c6655b more linting 2025-04-22 11:34:38 -07:00
Sydney RunkleandGitHub 44af4d8e3e Merge branch 'main' into sr/better-interrupts 2025-04-22 11:31:46 -07:00
Sydney Runkle 8e6c0be48a try linting + maybe test fix 2025-04-22 11:28:09 -07:00
Sydney Runkle 35523f4081 linting 2025-04-22 10:34:24 -07:00
Sydney Runkle 9a7b1fa12a initial pass - surfacing interrupts for stream_mode='values' 2025-04-22 10:15:19 -07:00
Nuno CamposandGitHub 71bf2f9e85 Rewrite graph drawing logic (#4354)
- It now executes the same pregel algo as when the graph is executed
(without running any user code in nodes or conditional edges) to
discover all the edges
- This means we now support drawing the graph for all Pregel instances,
not just StateGraph
- This is done in preparation for new edge/node type coming in separate
PR
- Known changes
  - custom labels on conditional edges to END are no longer displayed
2025-04-22 09:31:13 -07:00
Nuno Campos b03c647677 Lint 2025-04-22 08:45:50 -07:00
Nuno Campos 5d49188d3e Lint 2025-04-22 08:41:51 -07:00
Nuno Campos deeb2d6e92 Fix 2025-04-22 08:36:42 -07:00
Nuno Campos 0db67d4196 Fix 2025-04-22 08:36:42 -07:00
Nuno Campos 56c9c210c3 Fix 2025-04-22 08:36:42 -07:00
Nuno Campos 3398715258 Rewrite graph drawing logic
- It now executes the same pregel algo as when the graph is executed (without running any user code in nodes or conditional edges) to discover all the edges
- This means we now support drawing the graph for all Pregel instances, not just StateGraph
2025-04-22 08:36:41 -07:00
William Fu-Hinthorn 38d806733d Update site_description
Signed-off-by: William Fu-Hinthorn <13333726+hinthornw@users.noreply.github.com>
2025-04-22 08:18:10 -07:00
William FHandGitHub 86ddd8da10 Add docs on tunneling (#4371)
Signed-off-by: William Fu-Hinthorn <13333726+hinthornw@users.noreply.github.com>
2025-04-22 15:04:04 +00:00
William FHandGitHub a5f5d0c4df Expose --tunnel flag to dev command (#4370)
Signed-off-by: William Fu-Hinthorn <13333726+hinthornw@users.noreply.github.com>
cli==0.2.6
2025-04-22 14:23:09 +00:00
Sydney Runkle cba7d21732 fix tests? 2025-04-21 21:14:14 -07:00
Sydney Runkle b6ea73ff24 linting for 3.12 2025-04-21 21:01:27 -07:00
Sydney Runkle 1a477e57ff upgrade to py39 standards 2025-04-21 20:43:40 -07:00
lc-arjunandGitHub 7486adabdf fix: threads search sorting defaults (#4365)
Removes default from https://github.com/langchain-ai/langgraph/pull/4362
sdk==0.1.63
2025-04-21 20:49:16 -04:00
William FHandGitHub 12ad47e4e8 Use model_validate if needed (#4363)
If the state schema uses validators, skip the model construct
optimization.

For context, pydantic state can be significantly slower to run than
typed dict and dataclass states due to the full recursive validation.

We have some optimizations to reduce the impact of this (using cached
validators with model_construct), but this doesn't handle things like
field_validator.

We prefer correctness over performance, obviously.

Resolves: https://github.com/langchain-ai/langgraph/issues/4074

Signed-off-by: William Fu-Hinthorn <13333726+hinthornw@users.noreply.github.com>
2025-04-21 21:53:58 +00:00