LangChain

LangChain walks through the SRE agent running its own Kubernetes cluster: a scheduler gathers raw state through the Python client at zero token cost, one forced-tool Haiku call turns it into a severity-sorted Slack report, and specialist subagents fan out for diagnosis while all writes stay behind a

LangChain details building its own paid-media marketing agent on Deep Agents: treating the agent as a knowledge worker with a sandboxed computer, using code for calculations and models for judgment, which grew paid pipeline from 0-20% and cut reporting costs 40x.

Agent skills that teach LangChain concepts inside the coding agent you already use.

OpenWiki, LangChain's open-source agent CLI for writing and maintaining repo documentation, adds a web visualizer where humans can read the generated wikis and inspect how repo sections relate on a graph.

LangChain's Managed Deep Agents enters public beta, offering a managed path from prototype to production scale while keeping control over model choice and the agent development lifecycle.

A 17-minute tour of OpenWiki, LangChain's open-source self-maintaining wiki: how it works, why documentation built for agents looks nothing like documentation for humans, and the lessons from building it.

LangChain's Managed Deep Agents, built on the Harbor framework, runs every eval in a fresh container and logs results in LangSmith, aiming to catch regressions when you swap models, add skills, or edit tool descriptions in an agent.

LangSmith usage data shows open-weight models going mainstream: one in three active teams ran one in July, up from one in five, with daily open-weight usage growing 9x year over year against 4.8x for closed models.

WikiBench scores generated codebase wikis by having a reader agent answer repo-grounded questions, with separate judges for fact coverage and grounding. Across harnesses the spread is wide — $0.44 to $9.18, 11 to 50 minutes — and extra spend buys reading, not more pages.

An airline found 13% of messages to its concierge agent were classified as out of scope, and a review of production traces showed 95% of those were real passenger needs. Adding a customer care specialist cut the out of scope rate to 1% and lifted the return rate.

LangChain's MCP integration picks up the updated spec's stateless transport, along with elicitation support. The linked writeup also puts numbers on MCP adoption: Tier 1 SDKs approaching half a billion monthly downloads, and MCP tool calls from ChatGPT up 98x across 2026.

LangSmith's LLM Gateway adds data protection controls: PII and secrets detected, redacted or replaced before requests reach model providers, and scrubbed from recorded traces.

LangSmith's LLM Gateway adds model fallbacks: define rules across models and hosts once, and calls reroute automatically when a provider goes down or rate-limits.

OpenWiki 0.4.0 adds coding-agent integrations: init and update can be driven from Claude, Codex or OpenCode with almost no configuration, folding wiki generation into the agent loop instead of a standalone CLI run.

LangChain's agent issue-detection service ships an update: self-hosted deployment alongside SaaS, a cheaper Reduced Analysis mode, Slack and Linear integrations, automatic closing of stale issues, and a claimed 2x gain on internal detection benchmarks.

Similarweb combines deterministic tool-call checks, rubric-scored LLM judges, faithfulness checks and A/B comparisons against saved baselines, all connected to execution traces.

LangChain's Managed Deep Agents now provision the Slack app as part of deployment. No manifest to author, no OAuth redirect configuration, no bot tokens to copy between environments — one command and the agent shows up in the workspace.

OpenWiki 0.4.0 adds OKF v0.2 support and a claims runtime that keeps a persistent source of truth behind generated pages, surfacing which sources built a page, whether it is fully verified, and what produced it.

Slack's Add to Slack launch makes the messaging app a one-click deploy target for agents built in LangChain, OpenAI, n8n, Lovable, Vercel and other partner tools, skipping custom bot integration work.

LangChain's hosted Deep Agents now build their execution environment at deploy time from a setup.sh or Dockerfile, so each new thread starts on a prepared image instead of repeating clone and install work.

OpenAI's prompt cache cuts request cost by about 90%, but a single cache key tops out near 15 requests per second. Connor Heggie of Unify describes the routing layer they built across keys to hold a roughly 95% hit rate once traffic passed that ceiling.

Preview Builds are in public beta for LangSmith Deployment: a temporary production like deployment spun up from a pull request branch so agent changes can be traced and reviewed before merge, then torn down automatically afterwards.

Unify's Connor Heggie keeps the LLM judge in a different model family from the agent under test. A same-family pairing converges on shared blind spots, so the eval quietly approves the failure modes the judge cannot see either.

Rippling layers its agent evals four ways: offline fixtures with pre-recorded mocks on every commit, 300–400 sandbox queries after merge, roughly ten deploy-blocking scenarios against live systems, and continuous scheduled runs against production data.

LangChain has released Deep Agents Code (dcode), an open-source, model-agnostic coding agent, alongside a new LangChain Academy course covering installation and enterprise-oriented usage controls.

LangChain tests TypeSafe AI's Jev, a typed 'System One' classifier, against GPT-5.6 and Claude Sonnet LLM judges on a shared Deep Agents weather-task dataset. Jev showed 92-913x lower quality-score variance and ran at $0.00035/call versus $28 total for Claude, suggesting a new evaluator category for

LangChain revamped its MCP integration: support for the new stateless MCP spec, the integration folded into the main langchain package, elicitation delivered through interrupts, cacheable list endpoints, and a new adapter built on FastMCP. Expect import path and adapter changes.

LangChain and Nevermined published a cookbook for letting an agent transact mid-task without a human in the loop. It covers delegating payment credentials under explicit limits, having the agent buy credits and top itself up to pay providers, and tracing each purchase in LangSmith.

LangChain positions its Deep Agents framework around built-in context engineering: filesystems, subagents, and skills wired up specifically to keep long-running agent tasks from overloading the context window.

OpenWiki extended its coding agent integrations to Cursor, joining Claude Code, Codex and OpenCode. Setup is a global npm install of openwiki@latest followed by openwiki integrations install cursor.

LangChain describes Deep Life Sci, an open source agent harness built on deepagents that pulls from 600k+ clinical trial records and tens of millions of PubMed papers, using sub-agents and sandboxes for pharma R&D tasks.