Who contributed to AI Agents on August 19, 2026?

11 developers shipped this update, including github-actions[bot], langchain-oss-model-profiles[bot], ccurme, mdrxy, nazsats, hieptl, VascoSch92, and karan singh, and 3 more.

What were the notable AI Agents updates?

Desktop v0.0.14, fix: run_commands object form without args routes through the shell instead of failing with ENOENT (#13336), and langchain-openai==1.5.2.

AI Agents

LangChain, OpenHands, Cline, and LlamaIndex - agent frameworks and coding agents

Pick a date

Topics: AI Agents AI / ML Full archive →

The Wire · Showcase

CLINE DESKTOP GOES NATIVE WITH BACKGROUND ALERTS

By RepoJournal · Filed 07:02 UTC on August 19, 2026 · About AI Agents

11 people shipped this

github-actions[bot] @github-actions[bot] 2 cited

VascoSch92 @VascoSch92 2 cited

markmcd @markmcd 2 cited

nazsats @nazsats 1 cited

Saoud Rizwan @saoudrizwan 1 cited

langchain-oss-model-profiles[bot] @langchain-oss-model-profiles[bot]

ccurme @ccurme

mdrxy @mdrxy

hieptl @hieptl

karan singh @karan-963

JasmineLCY @JasmineLCY

Cline's desktop app just became a tool you can leave running while you grab coffee, and LangChain is quietly fixing the token math that makes DeepSeek caching actually count.

Cline shipped Desktop v0.0.14 [1], which finally posts native macOS notifications when a task finishes or needs attention, and it adds voice input with live transcription into the composer. Commands also stream their output into the transcript as they run instead of dumping everything at exit, so you can watch long tasks unfold without refreshing. On the backend, Cline fixed a session-breaking bug [2] where the structured run_commands form without args failed with ENOENT on any command containing a space, and it now routes skill slash commands through the skills tool instead of expanding them into your user message . LangChain released langchain-openai 1.5.2 [3], which among other things adds support for o-series models in token counting, and a fix [4] maps DeepSeek's "prompt_cache_hit_tokens" to the cache_read field so cache usage actually shows up in metadata. Meanwhile, OpenHands is batching StreamingDeltaEvents [5] so the UI can keep up with fast models, and it finished purging the old "microagent" terminology from the frontend [6]. LlamaIndex is shifting its Google GenAI integration to Gemini 3.7 Flash by default [7] and patching a bug where thought signatures could be overwritten [8]. There were 48 commits and 47 PRs across the four repos this period, and these two releases are the ones worth your attention.

Action items

→ Update Cline desktop to v0.0.14 and review the new notification and streaming settings cline/cline [plan]
→ Upgrade langchain-openai to 1.5.2 to get o-series token counting and the DeepSeek cache fix langchain-ai/langchain [plan]
→ Test OpenHands UI streaming if you rely on fast model output OpenHands/OpenHands [monitor]

References

[1] Desktop v0.0.14 ↗ cline/cline
[2] fix: run_commands object form without args routes through the shell instead of failing with ENOENT (#13336) cline/cline
[3] langchain-openai==1.5.2 ↗ langchain-ai/langchain
[4] fix(deepseek): map `prompt_cache_hit_tokens` to `cache_read` ↗ langchain-ai/langchain
[5] perf: batch StreamingDeltaEvents so the UI keeps up with fast models ↗ OpenHands/OpenHands
[6] refactor: finish the microagent→skill rename in the frontend ↗ OpenHands/OpenHands
[7] feat(google-genai): use gemini 3.7 Flash by default in lib and docs (#22734) run-llama/llama_index
[8] fix: dont overwrite thought signatures ↗ run-llama/llama_index

Quick answers

What shipped in AI Agents on August 19, 2026?: Cline's desktop app just became a tool you can leave running while you grab coffee, and LangChain is quietly fixing the token math that makes DeepSeek caching actually count. In total, 48 commits, 47 pull requests, and 2 releases landed.
Who contributed to AI Agents on August 19, 2026?: 11 developers shipped this update, including github-actions[bot], langchain-oss-model-profiles[bot], ccurme, mdrxy, nazsats, hieptl, VascoSch92, and karan singh, and 3 more.
What were the notable AI Agents updates?: Desktop v0.0.14, fix: run_commands object form without args routes through the shell instead of failing with ENOENT (#13336), and langchain-openai==1.5.2.

LangChain ships core 1.6.0 with a security warning tucked inside

LangChain's newest core release fixes a tool-injection bug, but the fix itself widens the path for arbitrary code execution.

ai-agents 95 shipped

AI Agents Aug 18

OpenHands ships Git Sync UI as LangChain rides the gateway wave

OpenHands finally put a face on its git sync automation, while LangChain's latest release pack quietly re-tools how agents talk to OpenAI.

ai-agents 69 shipped

AI Agents Aug 17

LANGCHAIN REFRESHES MODEL PROFILES AS OPENHANDS HARDENS LLM VALIDATION

LangChain's automated model profile system added DeepSeek-V4-Pro support while OpenHands deployed pre-flight LLM validation to catch misconfigured profiles before they ship.

ai-agents 8 shipped

AI Agents Aug 16

OPENHANDS PINS CLIENT DEPENDENCY, ADDS STRUCTURED ERROR HANDLING

OpenHands locked its TypeScript client to exact version 1.37.0 and merged canvas error outcome improvements in a single commit cycle.

ai-agents 2 shipped

Elsewhere on the wire

Local LLMs Aug 20

OLLAMA FIXES DEADLOCK THAT HUNG CHAT AND GENERATE

A mid-stream parser error in Ollama's server silently wedged chat and generate requests, leaking goroutines and hanging retries until a fix landed overnight.

+14

llms 268 shipped

@CachyOS Aug 20

LINUX-CACHYOS SHIPS 7.2.0-1

CachyOS's custom kernel just advanced to 7.2.0-1, and the sched-ext toolchain is moving in lockstep.

infra 55 shipped

Elixir & Phoenix Aug 20

LIVE VIEW HOOKS GET THE ELOQUENT TREATMENT

Phoenix Live View is shipping a passel of hook and JS-client fixes today, headlined by new beforeUpdate arguments that hand you more control over a hook's lifespan.