AICoder

AI Coding Live News

First-hand AI coding updates from 𝕏/Twitter frontier labs and researchers, distilled by AI with entity benchmarks.

Explore live on 𝕏 ↗

AI Coding Live News

C
Claude Code Changelog@ClaudeCodeLog·
👀112
🔥86
View on 𝕏 ↗

Claude Code 2.1.299 adds idle compaction for 200K+ token sessions before cache expiry and fixes sandboxed writes into --plugin-dir

Anthropic released Claude Code 2.1.299 on Oct 14. A new default compacts a 200K+ token terminal conversation after about 54 idle minutes, before its one-hour prompt cache expires, so the next message doesn't reprocess everything; tune it with CLAUDE_CODE_IDLE_COMPACT_MIN_TOKENS or turn it off with "idleCompaction": false. The release also fixes sandboxed shell commands being able to write into a --plugin-dir folder whose code Claude Code reloads outside the sandbox, and adds --disallow-tools to claude plugin eval.

Claude Code 2.1.299 adds idle compaction for 200K+ token sessions before cache expiry and fixes sandboxed writes into --plugin-dir

Key Takeaways

  • Idle compaction: triggers at 200K+ tokens after ~54 idle minutes, before the one-hour prompt cache expires
  • Configurable: CLAUDE_CODE_IDLE_COMPACT_MIN_TOKENS changes the 200K floor; "idleCompaction": false in settings.json disables it
  • Security fix: sandboxed commands could write into a --plugin-dir folder whose code is reloaded and run outside the sandbox
  • Custom ANTHROPIC_BASE_URL streams now wait 300 s instead of 180 s with no data; gateway-specific model names no longer fail compaction near 200K tokens
  • claude plugin eval gains --disallow-tools and a disallowed_tools case field to keep built-ins such as Agent out of eval runs
Read full story
Linked Models & Agents:
🤖 Claude Code🌐 Anthropic
#Claude Code#Anthropic#CLI#prompt caching#Compaction#Sandbox#Security#Plugins
L
LangChain@langchain·
👀77
🔥88

LangChain.js @langchain/openai 1.6.0: explicit prompt cache options + content-block breakpoints

LangChain.js shipped @langchain/[email protected] on 2026-09-27: ports explicit prompt-cache options across Chat Completions and Responses, preserves content-block breakpoints, and maps cache-write usage to cache_creation. A same-tag fix serializes promptCacheRetention: "in-memory" as API-legal "in_memory" and unifies precedence (per-call → modelKwargs → constructor); Chat Completions no longer drops modelKwargs.

LangChain.js @langchain/openai 1.6.0: explicit prompt cache options + content-block breakpoints

Key Takeaways

  • Shipped: @langchain/[email protected] / npm 1.6.0
  • Explicit prompt cache options + content-block breakpoints; cache-write → cache_creation (PR #11232)
  • Fix: serialize in-memory as in_memory; precedence per-call > modelKwargs > constructor (PR #11735)
  • Install: npm i @langchain/[email protected]
  • Follow-on: @langchain/[email protected] bumps this dependency the same day
Read full story
Linked Models & Agents:
⚡ GPT-6⚡ OpenAI🤖 LangChain.js🌐 LangChain🌐 OpenAI
#LangChain.js#OpenAI#Prompt cache#Langchain#TypeScript#cache_creation#AI-SDK-adjacent
M
Mastra@mastra_ai·
👀54
🔥90

Mastra @mastra/core 1.70.0: ModelSelectionProcessor routes cheap vs strong models; durable streams gain closeOnSuspend

Mastra published @mastra/[email protected] on 2026-09-24 with ModelSelectionProcessor for classifier-backed per-request model routing (#24865), closeOnSuspend on durable agent streams so AG-UI-style loops can finish on tool suspend, plus createClassifierScorer(), optional tool titles, and trace aggregate dimension/measure helpers with tag predicates. Docs warn routing is not always cheaper because models do not share prompt caches.

Key Takeaways

  • ModelSelectionProcessor routes simple vs hard requests (#24865); scope: 'first-step' optional
  • Durable closeOnSuspend: true ends streams at tool suspend (#24894)
  • createClassifierScorer() maps Classifier questions to 0–1 scores (#24792)
  • Trace aggregate descriptors + tag predicates for advanced queries
  • Upgrade: npm i @mastra/[email protected]; see CHANGELOG + processors docs
Read full story
Linked Models & Agents:
🤖 Mastra🌐 Mastra
#Mastra#TypeScript#ModelSelectionProcessor#durable agents#Classifiers#Observability
@
@AnthropicAI@AnthropicAI·
👀152
🔥88
View on 𝕏 ↗

Anthropic launches Claude Opus 5.5: kicks off 5.5 family with 40% cost reduction and 30%+ speedup

Anthropic officially launched Claude Opus 5.5 as the frontier pioneer of the new Claude 5.5 family. Benchmark audits show Opus 5.5 matches previous flagship Claude Fable 5.1 on core agentic coding and reasoning tasks while slicing operating costs by 40% compared to Opus 5. API pricing is set at $4/M input and $20/M output with $0.20/M prompt cache reads.

Key Takeaways

  • Flagship launch kicking off Claude 5.5 family, matching Claude Fable 5.1 performance on primary workflows.
  • Operating costs slashed by ~40% with over 30% faster token output speed.
  • API pricing at $4/M input, $20/M output, and $0.20/M prompt cache reads ($5/M cache write).
  • Highest automated behavioral safety audit score recorded by Anthropic to date.
Read full story
Linked Models & Agents:
#AI#LLM#Coding
ADSponsored
v
vechen@miu21590·
👀86
🔥88
View on 𝕏 ↗

Mid-task Jev in Codex raises/lowers GPT-6 Astra reasoning effort; author reports ~50% Astra cost cut without breaking prompt cache

@miu21590 demoed using Jev inside a running Codex session to raise/lower GPT-6 Astra reasoning effort—more thinking when stuck, less on routine steps—reporting ~50% lower Astra costs and faster runs without breaking prompt caching. Versus the usual pre-task Jev model pick, this wires the decision model mid-run for cost control.

Mid-task Jev in Codex raises/lowers GPT-6 Astra reasoning effort; author reports ~50% Astra cost cut without breaking prompt cache

Key Takeaways

  • Call Jev mid-Codex-run to raise/lower GPT-6 Astra reasoning effort.
  • Author tests: ~50% lower Astra cost, faster, prompt cache intact.
  • Moves Jev from pre-task model pick to in-task cost control.
Read full story
Linked Models & Agents:
🤖 codex🌐 OpenAI
#Codex#Jev#GPT-6#Astra#reasoning effort#prompt cache#Cost
吉
吉岡 裕貴@yoppy0123·
👀42
🔥88
View on 𝕏 ↗

Claude Code v2.1.260: Fable 5.1 keeps prompt cache valid when effort changes mid-chat

Developer Yuki Yoshioka (@yoppy0123) notes that from Claude Code v2.1.260, Fable 5.1 keeps the prompt cache valid when effort is changed mid-conversation, cutting repeated prefill cost from effort switches. The post warns Opus 5 is not listed in the notes and likely lacks the same fix yet, so treat models differently after upgrading.

Key Takeaways

  • Claude Code v2.1.260: Fable 5.1 mid-chat effort changes no longer invalidate the prompt cache.
  • Helps adjust reasoning effort without re-paying large context prefill.
  • Opus 5 not confirmed in notes — do not assume all models behave the same.
Read full story
Linked Models & Agents:
⚡ Claude Fable 5.1⚡ Claude Opus 5🤖 Claude Code (Anthropic)🌐 Anthropic
#Claude Code#Fable 5.1#prompt cache#Effort#Opus 5#Anthropic
You've caught up with all stories