Skip to content
Problems.

Anthropic upgrades break working setups

Outages and change7 posts from 6 people7d active+11 post in the last 7 days, 2 the 7 days before (steady)

Posts per day

Posts per day7 posts, Sep 3 to Sep 25

The posts behind it

7, newest first
PostDate
core: empty `group_ids` bypasses validation and filtering in `InMemoryRecordManager`### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...langchain-ai/langchainNikhi00718Sep 223 days ago
UsageMetadataCallbackHandler silently drops usage_metadata when model_name is missing### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...langchain-ai/langchainammar-iitmSep 169 days ago
Token budget middleware that checks in with the user instead of hard-stopping...as tool names or a predicate, so tools like `ask_user` or `save_progress` are never blocked by the budget and calling them does not end the run. - **Safety net.** If the model still calls a restricted tool after the check-in, the call is executed and the next `before_model` ends...langchain-ai/langchainsauravsolankiSep 1411 days ago
BaseLLM cache ignores per-call generation options and replays the wrong completion### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...langchain-ai/langchainyunfeizhuSep 112 weeks ago
500 "no user query found in messages" for tool-only turns on /v1/messages and /v1/chat/completions (no truncation required)...through the model template like any other turn. Requiring a literal user text turn and failing with 500 otherwise breaks spec-compliant agent clients (Anthropic/OpenAI SDKs) whenever they send a follow-up turn that only returns tool output -- and permanently so once truncation...ollama/ollamajolestarSep 82 weeks ago
ContextEditingMiddleware blocks the event loop with model-based token counting### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...langchain-ai/langchainChampionsZhangSep 72 weeks ago
`HumanInTheLoopMiddleware`: cross-tool edit bypasses the target tool's `interrupt_on` policy### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...langchain-ai/langchainChampionsZhangSep 33 weeks ago

Companies and products named

Company or productPosts naming it
Anthropic7
OpenAI7
Ollama7
Perplexity6
DeepSeek6
Hugging Face6
About this problem

Evidence

7 posts from 6 people in 2 places, about 2 a week over 20 days. Mostly on langchain-ai/langchain, ollama/ollama. Tools named alongside: OpenAI, Ollama, Perplexity, DeepSeek.

Frustration Frustration 0 of 3· Seen on GitHub issues

How it was grouped

Posts that state a pain and match the "breaking-changes" rule about Anthropic. First post Sep 3, 2026, latest Sep 22, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method

History

  • 2026-09-25 Added: Anthropic upgrades break working setups (7 posts)

Rising problems by email

Mondays: the problems in data, tech and AI that grew fastest that week.

Double opt-in. Unsubscribe any time.