# Talking Shop **URL:** https://forum.langchain.com/c/talking-shop/12.md?page=1 [Latest](https://forum.langchain.com/latest.md) · [Categories](https://forum.langchain.com/categories.md) · [Tags](https://forum.langchain.com/tags.md) **Page:** 2 --- ## [How do you actually decide between LangGraph, AutoGen, and which model to use for a task?](https://forum.langchain.com/t/how-do-you-actually-decide-between-langgraph-autogen-and-which-model-to-use-for-a-task/4251)
**Author:** [@aditijain451](https://forum.langchain.com/u/aditijain451)\ **Replies:** 0\ **Last updated:** [July 27, 2026, 11:41am UTC](https://forum.langchain.com/t/how-do-you-actually-decide-between-langgraph-autogen-and-which-model-to-use-for-a-task/4251 "2026-07-27T11:41:32Z")
Been building agent workflows for a bit now, and every time I need to pick a framework + model combo, it feels like I’m going on vibes — a blog post, whatever worked last time, whatever’s trending on Twitter that week. S… --- ## [The token cost of long agent runs compounds harder than I expected — I measured 66 real sessions](https://forum.langchain.com/t/the-token-cost-of-long-agent-runs-compounds-harder-than-i-expected-i-measured-66-real-sessions/4240)
**Author:** [@reelwiselab](https://forum.langchain.com/u/reelwiselab)\ **Replies:** 1\ **Last updated:** [July 27, 2026, 3:22am UTC](https://forum.langchain.com/t/the-token-cost-of-long-agent-runs-compounds-harder-than-i-expected-i-measured-66-real-sessions/4240 "2026-07-27T03:22:14Z")
Full disclosure up front: I run an autonomous AI dev studio, and one thing that kept surprising us was how quickly long agent runs get expensive. So I instrumented 66 real coding-agent sessions and broke the spend down b… --- ## [I built AgentPulse to catch silent drift in LangChain and LangGraph workflows](https://forum.langchain.com/t/i-built-agentpulse-to-catch-silent-drift-in-langchain-and-langgraph-workflows/4233)
**Author:** [@leyla0016](https://forum.langchain.com/u/leyla0016)\ **Replies:** 0\ **Last updated:** [July 24, 2026, 1:07pm UTC](https://forum.langchain.com/t/i-built-agentpulse-to-catch-silent-drift-in-langchain-and-langgraph-workflows/4233 "2026-07-24T13:07:46Z")
I’ve been working on AgentPulse, an open-source, self-hosted tool for detecting and investigating behavioral drift in multi-agent systems. LangChain and LangGraph felt like a natural fit because AgentPulse works with th… --- ## [Is LangChain killing its own Innovation with gatekeeping?](https://forum.langchain.com/t/is-langchain-killing-its-own-innovation-with-gatekeeping/4069)
**Author:** [@leo-gan](https://forum.langchain.com/u/leo-gan)\ **Replies:** 5\ **Last updated:** [July 21, 2026, 4:44pm UTC](https://forum.langchain.com/t/is-langchain-killing-its-own-innovation-with-gatekeeping/4069 "2026-07-21T16:44:25Z")
The LangChain community used to be vibrant. Anyone could add a feature or fix a bug, and pull requests were often merged on the same day. Contributions were welcomed, which helped LangChain grow in popularity; it was eas… --- ## [Do your tool-using agents re-plan the same task every single time ?](https://forum.langchain.com/t/do-your-tool-using-agents-re-plan-the-same-task-every-single-time/4180)
**Author:** [@siddhu1716](https://forum.langchain.com/u/siddhu1716)\ **Replies:** 2\ **Last updated:** [July 17, 2026, 1:30pm UTC](https://forum.langchain.com/t/do-your-tool-using-agents-re-plan-the-same-task-every-single-time/4180 "2026-07-17T13:30:22Z")
I’ve noticed something while building tool-using agents (ReAct, LangGraph, OpenAI tool calling). Even when the task has essentially the same structure—say: fetch active users filter by condition export CSV t… --- ## [How are people making multi-step LLM workflows inspectable and rerunnable?](https://forum.langchain.com/t/how-are-people-making-multi-step-llm-workflows-inspectable-and-rerunnable/4172)
**Author:** [@Levitan-factory](https://forum.langchain.com/u/Levitan-factory)\ **Replies:** 2\ **Last updated:** [July 16, 2026, 9:06pm UTC](https://forum.langchain.com/t/how-are-people-making-multi-step-llm-workflows-inspectable-and-rerunnable/4172 "2026-07-16T21:06:05Z")
I keep running into the same problem with LLM workflows: the useful part is no longer just the prompt, but the trail around it. For one-off exploration, chat is fine. For anything repeatable, I want to know: which mod… --- ## [Adaptive Explanation, Conditional Plans, and Plan Reuse](https://forum.langchain.com/t/adaptive-explanation-conditional-plans-and-plan-reuse/3959)
**Author:** [@AdamSobieski](https://forum.langchain.com/u/AdamSobieski)\ **Replies:** 2\ **Last updated:** [July 4, 2026, 10:21pm UTC](https://forum.langchain.com/t/adaptive-explanation-conditional-plans-and-plan-reuse/3959 "2026-07-04T22:21:32Z")
Hello. For the use case of adaptive explanation, answering users’ questions and participating in explanatory dialogues with users while making use of dynamic user models, I am exploring conditional planning, where agents… --- ## [Mapping OWASP ASI06 (memory poisoning) flows in a LangGraph email agent — what I found](https://forum.langchain.com/t/mapping-owasp-asi06-memory-poisoning-flows-in-a-langgraph-email-agent-what-i-found/4079)
**Author:** [@Arul](https://forum.langchain.com/u/Arul)\ **Replies:** 0\ **Last updated:** [July 2, 2026, 5:18pm UTC](https://forum.langchain.com/t/mapping-owasp-asi06-memory-poisoning-flows-in-a-langgraph-email-agent-what-i-found/4079 "2026-07-02T17:18:17Z")
I’ve been building a static, repo-level scanner for OWASP ASI06 (memory & context poisoning) in agent code, and I wanted to test it against a real LangGraph agent rather than my own. I ran it on a small public LangGraph … --- ## [Seeking Feedback & Collaboration: Talordata - A Multi-Engine SERP API for LangChain Agents & RAG Workflows](https://forum.langchain.com/t/seeking-feedback-collaboration-talordata-a-multi-engine-serp-api-for-langchain-agents-rag-workflows/4065)
**Author:** [@TensorVoyager](https://forum.langchain.com/u/TensorVoyager)\ **Replies:** 0\ **Last updated:** [June 30, 2026, 1:26pm UTC](https://forum.langchain.com/t/seeking-feedback-collaboration-talordata-a-multi-engine-serp-api-for-langchain-agents-rag-workflows/4065 "2026-06-30T13:26:45Z")
Hi everyone, I’ve been working with LangChain (and LangGraph) for building agents and RAG systems. One pain point I often encounter is the cost and single-engine limitation of existing search tools (like Tavily, Serper,… --- ## [tracesage: See Inside Your LangGraph Agents](https://forum.langchain.com/t/tracesage-see-inside-your-langgraph-agents/4063)
**Author:** [@kjgpta](https://forum.langchain.com/u/kjgpta)\ **Replies:** 0\ **Last updated:** [June 30, 2026, 1:26pm UTC](https://forum.langchain.com/t/tracesage-see-inside-your-langgraph-agents/4063 "2026-06-30T13:26:34Z")
As LangGraph applications grow from simple prototypes into production systems, debugging becomes increasingly difficult. A single user request may involve: Multiple agents Tool calls Retrieval pipelines MCP servers Se… --- ## [Issue with thinking\_level Parameter in ChatGoogleGenerativeAI (langchain-google-genai v3.1.0)](https://forum.langchain.com/t/issue-with-thinking-level-parameter-in-chatgooglegenerativeai-langchain-google-genai-v3-1-0/2291)
**Author:** [@MihoKey](https://forum.langchain.com/u/MihoKey)\ **Replies:** 5\ **Last updated:** [June 29, 2026, 9:28pm UTC](https://forum.langchain.com/t/issue-with-thinking-level-parameter-in-chatgooglegenerativeai-langchain-google-genai-v3-1-0/2291 "2026-06-29T21:28:43Z")
I am encountering a ValueError while attempting to utilize the new thinking\_level parameter in ChatGoogleGenerativeAI after upgrading to langchain-google-genai version 3.1.0. The inclusion of this field in the configura… --- ## [Post-retrieval temporal decay : how are you handling stale context in production RAG pipelines?](https://forum.langchain.com/t/post-retrieval-temporal-decay-how-are-you-handling-stale-context-in-production-rag-pipelines/3968)
**Author:** [@VLSiddarth](https://forum.langchain.com/u/VLSiddarth)\ **Replies:** 2\ **Last updated:** [June 27, 2026, 7:18am UTC](https://forum.langchain.com/t/post-retrieval-temporal-decay-how-are-you-handling-stale-context-in-production-rag-pipelines/3968 "2026-06-27T07:18:32Z")
We’ve been running into a specific failure mode in production RAG agents that I haven’t seen discussed much here: temporal staleness at the retrieval layer. The problem: standard vector search returns documents by se… --- ## [Exploring a LangChain + real-time voice integration pattern](https://forum.langchain.com/t/exploring-a-langchain-real-time-voice-integration-pattern/4039)
**Author:** [@bluemotional](https://forum.langchain.com/u/bluemotional)\ **Replies:** 0\ **Last updated:** [June 26, 2026, 4:47pm UTC](https://forum.langchain.com/t/exploring-a-langchain-real-time-voice-integration-pattern/4039 "2026-06-26T16:47:12Z")
Hi folks, I’m pretty new to LangChain myself. Before working on this, I honestly didn’t know much about it. As I started exploring it, I found myself wondering whether there are some cases where LangChain-style apps wo… --- ## [Show & Tell: Orchid – Local-first record and replay for AI agent debugging](https://forum.langchain.com/t/show-tell-orchid-local-first-record-and-replay-for-ai-agent-debugging/4025)
**Author:** [@marioguerra](https://forum.langchain.com/u/marioguerra)\ **Replies:** 0\ **Last updated:** [June 24, 2026, 3:56pm UTC](https://forum.langchain.com/t/show-tell-orchid-local-first-record-and-replay-for-ai-agent-debugging/4025 "2026-06-24T15:56:31Z")
Hey everyone, just released Orchid, (Orchestration interactive debugger) a zero-instrumentation proxy that captures every API & LLM call in your agentic pipeline, then lets you inspect and replay the entire run locally, … --- ## [A suggestion regarding the documentation](https://forum.langchain.com/t/a-suggestion-regarding-the-documentation/4016)
**Author:** [@Armagdy](https://forum.langchain.com/u/Armagdy)\ **Replies:** 1\ **Last updated:** [June 23, 2026, 9:23am UTC](https://forum.langchain.com/t/a-suggestion-regarding-the-documentation/4016 "2026-06-23T09:23:04Z")
Hi LangChain team, please avoid relying too heavily on AI-generated content without thorough human review. Some pages read well but miss important implementation details and proper explanations, making it difficult to b… --- ## [What does the security architecture of AI agents actually look like?](https://forum.langchain.com/t/what-does-the-security-architecture-of-ai-agents-actually-look-like/3105)
**Author:** [@joy7758](https://forum.langchain.com/u/joy7758)\ **Replies:** 11\ **Last updated:** [June 21, 2026, 10:30am UTC](https://forum.langchain.com/t/what-does-the-security-architecture-of-ai-agents-actually-look-like/3105 "2026-06-21T10:30:51Z")
As agent systems move into real workflows, security concerns go beyond simple prompt guardrails. From what I’ve been observing, agent security tends to involve several layers around the orchestration framework: 1 Run… --- ## [Open source project share : a local agent with safe document edits and dynamic direct return](https://forum.langchain.com/t/open-source-project-share-a-local-agent-with-safe-document-edits-and-dynamic-direct-return/3984)
**Author:** [@eric-burel](https://forum.langchain.com/u/eric-burel)\ **Replies:** 0\ **Last updated:** [June 19, 2026, 4:29pm UTC](https://forum.langchain.com/t/open-source-project-share-a-local-agent-with-safe-document-edits-and-dynamic-direct-return/3984 "2026-06-19T16:29:03Z")
Hey folks, I just wanted to share an open source agent I am building with LangChain, called “fais” : GitHub - lbke/fais: Fais - It does what you tell it to do · GitHub It’s meant to run as a CLI and focuses on administ… --- ## [Recommended migration path for tools and integrations following the langchain-community sunset?](https://forum.langchain.com/t/recommended-migration-path-for-tools-and-integrations-following-the-langchain-community-sunset/3934)
**Author:** [@VaibhavBhinge](https://forum.langchain.com/u/VaibhavBhinge)\ **Replies:** 1\ **Last updated:** [June 17, 2026, 6:03am UTC](https://forum.langchain.com/t/recommended-migration-path-for-tools-and-integrations-following-the-langchain-community-sunset/3934 "2026-06-17T06:03:54Z")
Hi everyone, With the recent official announcement sunsetting the langchain-community package and archiving the repository, I am evaluating how to safely migrate my current production workflows and agent toolsets. Prev… --- ## [Governance/policy checkpoint node pattern for LangGraph agents](https://forum.langchain.com/t/governance-policy-checkpoint-node-pattern-for-langgraph-agents/3791)
**Author:** [@nagasatish](https://forum.langchain.com/u/nagasatish)\ **Replies:** 5\ **Last updated:** [June 16, 2026, 4:24am UTC](https://forum.langchain.com/t/governance-policy-checkpoint-node-pattern-for-langgraph-agents/3791 "2026-06-16T04:24:34Z")
Hey all — I’ve been building production LangGraph agents and kept running into the same pattern: needing policy enforcement between the “decide” and “act” steps. Thought I’d share the pattern and see if others find it us… --- ## [What is wrong with Langchain frontend sdk](https://forum.langchain.com/t/what-is-wrong-with-langchain-frontend-sdk/3914)
**Author:** [@AmgadEbaid](https://forum.langchain.com/u/AmgadEbaid)\ **Replies:** 15\ **Last updated:** [June 16, 2026, 12:33am UTC](https://forum.langchain.com/t/what-is-wrong-with-langchain-frontend-sdk/3914 "2026-06-16T00:33:26Z")
I have a backend with custome endpoint that returns the correct sse format for langchain agent And when i use this endpoint with langchain frontend sdk with react or next js I noticed that it loses all the previous con… --- ## [Building a Regulator-Grade AI Runtime on LangGraph — Open Collaboration](https://forum.langchain.com/t/building-a-regulator-grade-ai-runtime-on-langgraph-open-collaboration/2041)
**Author:** [@eduardobb96](https://forum.langchain.com/u/eduardobb96)\ **Replies:** 1\ **Last updated:** [June 15, 2026, 2:15pm UTC](https://forum.langchain.com/t/building-a-regulator-grade-ai-runtime-on-langgraph-open-collaboration/2041 "2026-06-15T14:15:30Z")
Hey everyone :waving\_hand: We’re exploring LangGraph as part of a secure AI orchestration system called Aura — an on-prem, compliance-ready runtime designed for enterprises (currently piloted with banks). The core idea… --- ## [How to Make Your LangChain Agent EU AI Act Compliant in 5 Minutes](https://forum.langchain.com/t/how-to-make-your-langchain-agent-eu-ai-act-compliant-in-5-minutes/3228)
**Author:** [@AgentSnipper](https://forum.langchain.com/u/AgentSnipper)\ **Replies:** 1\ **Last updated:** [June 15, 2026, 2:11pm UTC](https://forum.langchain.com/t/how-to-make-your-langchain-agent-eu-ai-act-compliant-in-5-minutes/3228 "2026-06-15T14:11:55Z")
The EU AI Act requires human oversight (Article 14), audit logging (Article 12), and risk management (Article 9) for production AI agents. Most LangChain deployments have none of these. Here’s how to add all three in und… --- ## [Show LangChain: Catching silent memory leaks in agentic zero-copy swarms](https://forum.langchain.com/t/show-langchain-catching-silent-memory-leaks-in-agentic-zero-copy-swarms/3948)
**Author:** [@msunda17](https://forum.langchain.com/u/msunda17)\ **Replies:** 0\ **Last updated:** [June 15, 2026, 1:35pm UTC](https://forum.langchain.com/t/show-langchain-catching-silent-memory-leaks-in-agentic-zero-copy-swarms/3948 "2026-06-15T13:35:45Z")
LLMs fail continuous memory math 65% of the time. And standard unit tests don’t catch it. If you are building zero-copy LangGraph swarms or using shared prefix caching to reduce API costs across multiple agents, your LL… --- ## [Adding merchant verification to AI shopping agents with MCP](https://forum.langchain.com/t/adding-merchant-verification-to-ai-shopping-agents-with-mcp/3946)
**Author:** [@warwickwood-cell](https://forum.langchain.com/u/warwickwood-cell)\ **Replies:** 0\ **Last updated:** [June 15, 2026, 1:34pm UTC](https://forum.langchain.com/t/adding-merchant-verification-to-ai-shopping-agents-with-mcp/3946 "2026-06-15T13:34:42Z")
We built a small LangGraph + MCP example showing how AI shopping agents can verify ecommerce merchants before making recommendations or transaction decisions. The problem nobody is talking about: There is currently no … --- ## [LangChain/LangGraph tool args validation middleware](https://forum.langchain.com/t/langchain-langgraph-tool-args-validation-middleware/3910)
**Author:** [@Serjbory](https://forum.langchain.com/u/Serjbory)\ **Replies:** 3\ **Last updated:** [June 12, 2026, 7:16pm UTC](https://forum.langchain.com/t/langchain-langgraph-tool-args-validation-middleware/3910 "2026-06-12T19:16:50Z")
Hello ! I would like to sharing a small middleware I’ve been using to harden agents against malformed tool calls. ToolArgsValidationMiddleware validates LLM-generated tool-call arguments against each tool’s schema b… --- ## [\[Feature\] : Add ArabicTextSplitter with proper Arabic language boundary handling](https://forum.langchain.com/t/feature-add-arabictextsplitter-with-proper-arabic-language-boundary-handling/3888)
**Author:** [@KMalek101](https://forum.langchain.com/u/KMalek101)\ **Replies:** 0\ **Last updated:** [June 7, 2026, 10:22pm UTC](https://forum.langchain.com/t/feature-add-arabictextsplitter-with-proper-arabic-language-boundary-handling/3888 "2026-06-07T22:22:23Z")
Current text splitters are designed for English and don’t handle Arabic-specific characters correctly — Arabic punctuation (،, ؟, ؛), kashida (ـ), and diacritics cause incorrect chunk boundaries in Arabic RAG pipelines. … --- ## [Best Practices for Dynamic User Modeling with LangGraph](https://forum.langchain.com/t/best-practices-for-dynamic-user-modeling-with-langgraph/3884)
**Author:** [@AdamSobieski](https://forum.langchain.com/u/AdamSobieski)\ **Replies:** 1\ **Last updated:** [June 7, 2026, 12:16pm UTC](https://forum.langchain.com/t/best-practices-for-dynamic-user-modeling-with-langgraph/3884 "2026-06-07T12:16:52Z")
Hello. I am new to the community and am interested in dynamic user modeling for use cases like adaptive explanation, the custom-tailoring of unfolding explanatory dialogues for individual users. I am wondering about bes… --- ## [Best Practices for Enforcing Tool Call Order in LangGraph?](https://forum.langchain.com/t/best-practices-for-enforcing-tool-call-order-in-langgraph/3862)
**Author:** [@Eniola](https://forum.langchain.com/u/Eniola)\ **Replies:** 3\ **Last updated:** [June 5, 2026, 2:09pm UTC](https://forum.langchain.com/t/best-practices-for-enforcing-tool-call-order-in-langgraph/3862 "2026-06-05T14:09:53Z")
I’m trying to enforce a deterministic tool-calling workflow in LangGraph and would appreciate guidance on best practices. For example, say I have three tools: list\_insurance\_plans – shows all available insurance plan… --- ## [3rd party rate limit handling(OpenAI-level rate limiting in LangChain.js)](https://forum.langchain.com/t/3rd-party-rate-limit-handling-openai-level-rate-limiting-in-langchain-js/3845)
**Author:** [@parshvJS](https://forum.langchain.com/u/parshvJS)\ **Replies:** 2\ **Last updated:** [June 2, 2026, 12:20pm UTC](https://forum.langchain.com/t/3rd-party-rate-limit-handling-openai-level-rate-limiting-in-langchain-js/3845 "2026-06-02T12:20:32Z")
\]I am having a problem implementing OpenAI-level rate limiting in LangChain.js. I have explored the LangChain documentation and found two types of rate limiting: Agent-level rate limiting via middleware(for controllin… --- ## [Making Langgraph agent states and memories inspecatble](https://forum.langchain.com/t/making-langgraph-agent-states-and-memories-inspecatble/3839)
**Author:** [@muhammad.abdullah](https://forum.langchain.com/u/muhammad.abdullah)\ **Replies:** 0\ **Last updated:** [June 2, 2026, 2:40am UTC](https://forum.langchain.com/t/making-langgraph-agent-states-and-memories-inspecatble/3839 "2026-06-02T02:40:52Z")
Often, during development and production of agentic, montoring traces themselves is not enough, especially when issues might lie in the state checkpointers or in the memory stores. I built an mcp called LangMCP that help… [Previous page](https://forum.langchain.com/c/talking-shop/12.md) [Next page](https://forum.langchain.com/c/talking-shop/12.md?page=2)