MindPattern
← The Wire
Source trail

Releasebot

Public MindPattern findings, entities, and graph evidence that cite this source.

Findings
40
All-time hits
156
High value
33
Last seen
2026-07-07

Connected entities

Claude Code 2.1.191 Lets /rewind Resume From Before a /clear and Makes Background-Agent Stops PermanAnthropic Ships Claude GA in Microsoft Foundry on Azure — Opus 4.8 and Haiku 4.5 in the Messages APIReleasebotOpenAICodex

Related findings

  1. 2026-07-07 / TOOLSTip: Cut Tool-Schema Token Bloat With MCP Tool-Search-by-Default and Live /mcp Status PanelsLoading every MCP tool's schema into context is a silent token tax; Codex now does tool search by default, fetching schemas on demand instead of front-loading all of them, which keeps compatibility with older models. When a tool won't invoke, use the live /mcp status panel (now in both Codex and Devin Desktop) to confirm which servers actually connected rather than guessing. Together they reduce context waste and debugging time on multi-server setups.
  2. 2026-07-07 / TOOLSZed July Update: Telescope-Style Pickers, Agent Auto-Compaction, and New OpenCode ModelsZed added Telescope-style resizable pickers with live file/text previews and shipped auto-compaction plus a /compact command in Zed Agent, with skills management moved into the settings UI. On the model side it added GLM 5.2 and Kimi K2.7 Code to OpenCode Go and made DeepSeek V4 Pro available on OpenCode Zen. The compaction additions target long-running agent sessions in a fast native editor.
  3. 2026-07-07 / TOOLSOpenAI Codex July Update: MCP Tool-Search by Default, Multi-Agent Delegation Controls, and Rollout Token BudgetsCodex's July release makes MCP tool search the default (improving discovery while preserving older-model compatibility) and adds configurable rollout token budgets that track usage across agent threads and abort a turn when exhausted. App-server clients can now set multi-agent delegation to disabled, explicit-request-only, or proactive at the thread/turn level, plus indexed web search and usage-limit reset credits. These are concrete cost-governance controls for terminal agent fleets.
  4. 2026-07-07 / TOOLSClaude Code v2.1.202 Adds Dynamic-Workflow Size Control and Switches Default Permission Mode to ManualThe July 6 v2.1.202 release adds a 'Dynamic workflow size' setting in /config (small/medium/large) that governs how many subagents Claude spawns per workflow, plus workflow.run_id and workflow.name OpenTelemetry attributes for tracing workflow-spawned agents. It also changes the default permission mode to 'Manual' across the CLI, --help, VS Code, and JetBrains, alongside crash, tmux-render, and screen-reader fixes. The workflow-size knob is the practical lever for capping fan-out cost.
  5. 2026-07-02 / SOURCESAnthropic Ships July Claude Code + Cowork Updates: Effort Control, Self-Hosted Sandboxes, Cowork ObservabilityAnthropic's early-July release notes add a user-facing 'effort control' selector in claude.ai and Cowork (choose how deeply Claude thinks per response), self-hosted sandboxes for Claude Managed Agents as an alternative to running tool execution on Anthropic infrastructure, and Cowork support for the Analytics API plus OpenTelemetry monitoring. Claude Code also gained sandbox credential blocking and org-level model restrictions. These are directly actionable for teams operationalizing agents — the observability and self-hosted-execution pieces in particular address enterprise control and auditability concerns.
  6. 2026-07-02 / TOOLSTip: Use Claude Code's Org Default Models, Readable Session Names, and Clickable AttachmentsThe latest Claude Code update adds org-level default model configuration, human-readable session names, clickable file attachments, and a smoother agents view. Teams can pin a default model org-wide (e.g., standardize on Sonnet 5) so individual configs don't drift, and readable session names make it far easier to reattach to the right background agent among many.
  7. 2026-07-02 / TOOLSClaude Code Background Agents Now Auto-Commit, Push, and Open a Draft PR on FinishBackground agents launched from `claude agents` now commit, push, and open a draft PR when they finish code work in a worktree, instead of stopping to ask for permission. This closes the last manual step in the fire-and-forget worktree loop, letting a builder dispatch parallel agents and review finished PRs rather than babysitting each completion prompt.
  8. 2026-07-02 / TOOLSClaude in Chrome Reaches General AvailabilityAnthropic moved Claude in Chrome to general availability on July 1, 2026, putting a browser-native agent surface alongside Claude Code's terminal and IDE surfaces. For builders, it consolidates 'browser-as-tool' into a first-party capability rather than a third-party MCP add-on, closing the loop between web research, form-driven tasks, and agent execution.
  9. 2026-07-01 / TOOLSClaude Apps Gateway Lands for Amazon Bedrock and Google CloudThe June 30 release introduces a self-hosted control plane for Claude Code with corporate SSO, centrally enforced policy, role-based access, per-user cost tracking, and spend caps, deployable on Amazon Bedrock and Google Cloud. This is the enterprise-governance layer that lets orgs run Claude Code centrally rather than per-developer. Distinct from the agent-orchestration bundle GA — this is about deployment control, not subagent features.
  10. 2026-06-30 / REDDITAnthropic Ships Claude GA in Microsoft Foundry on Azure — Opus 4.8 and Haiku 4.5 in the Messages APIOn June 29, Anthropic made Claude generally available inside Microsoft Foundry on Azure, exposing Claude Opus 4.8 and Haiku 4.5 through the Messages API with Azure-native identity, billing, governance, and optional US data-zone support. It offers a hosted-on-Azure path so enterprises can run Claude entirely within their Azure environment. This puts Anthropic's frontier and fast tiers directly in front of Azure-committed enterprises, a notable distribution win against OpenAI's home-turf cloud.
  11. 2026-06-29 / DISPATCHOpenAI Makes Codex Remote GA Across All Plans and Retires GPT-4.5 in ChatGPTAs of June 26, OpenAI removed GPT-4.5 from ChatGPT (migrating existing conversations to GPT-5.5) and rolled Codex Remote to general availability on every plan, letting users start or resume coding work on connected Mac or Windows hosts directly from the mobile app. It's a concrete change to the Codex agentic-coding workflow that affects daily practitioners.
  12. 2026-06-28 / SOURCESClaude Code June 27 Release: Fullscreen Mouse Controls, Exact-Match Hook Matchers, Non-Blocking Background AgentsDistinct from the June 25 update, this release adds fullscreen mouse click/drag/hover controls (with a CLAUDE_CODE_DISABLE_MOUSE_CLICKS escape hatch that keeps wheel scroll), and fixes hook matchers with hyphenated identifiers (e.g. code-reviewer, mcp__brave-search) that were accidentally substring-matching — they now exact-match. Background agents no longer instruct the model to 'end your response,' so the parent keeps working while an agent runs, and MCP auth now auto-reconnects on 401/403. Small but high-leverage quality-of-life fixes for daily Claude Code users running hooks and background agents.
  13. 2026-06-28 / TOOLSClaude Code v2.1.185–192: `claude mcp login`, /rewind Past a /clear, and Background-Agent Permission SurfacingThe week-26 (June 22–26) Claude Code train added `claude mcp login` for authenticating configured MCP servers directly from the shell, /rewind support for resuming a conversation from before /clear was run, shell mode that responds to command output, and background subagents that now surface permission prompts in the main session instead of silently auto-denying. A startup notice also flags when MCP servers need authentication. These land just under yesterday's v2.1.193 (shell-classifier routing + assistant-response telemetry).
  14. 2026-06-26 / REDDITClaude Code Ships Reliability Update: /rewind, MCP Resilience, OAuth, Lower CPU/MemoryAnthropic shipped a broad Claude Code quality-and-reliability update this week including /rewind support, improved agent and permissions behavior, better MCP resilience and OAuth handling, smarter sandbox prompts, and reduced CPU/memory use during streaming and long sessions. Incremental, but a meaningful quality-of-life pass for daily Claude Code users running long agent sessions.
  15. 2026-06-26 / HACKER NEWSAnthropic Ships 'Claude Design' Updates — Design-System Sync, Direct Canvas Editing, Tighter Claude Code HandoffClaude Design received updates that keep generated work aligned to a project's design system, add direct canvas editing and stronger layout controls, and tighten Claude Code sync for smoother design-to-code handoff plus more tool connections. The thrust is closing the gap between design intent and generated implementation. For design-literate builders, enforcing a design system at generation time is more useful than after-the-fact cleanup of off-spec output.
  16. 2026-06-26 / TOOLSClaude Code 2.1.191 Lets /rewind Resume From Before a /clear and Makes Background-Agent Stops PermanentThe June 24 release adds /rewind support for resuming a conversation from before /clear was run — so you can wipe context to start a new task, then jump back into the prior thread if you discover you still need it. It also fixes background agents resurrecting after being stopped (tasks-panel stops are now permanent), fixes scroll jumping to the bottom during streaming, and cuts streaming-response CPU usage by roughly 37%.
  17. 2026-06-25 / DISPATCHMistral Launches Enterprise 'Connectors' With Scoped API Keys and Announces a 10 MW Inference Facility at Les UlisAt the AI Now Summit (June 24), Mistral shipped new Connectors capabilities for secure enterprise integrations — richer admin controls, scoped API keys, multi-account connectors, a debugger, and connector support inside Vibe Code and Workflows. It also announced a dedicated 10 MW inference facility in Les Ulis (Essonne) opening Q3 2026. The pairing of secure agent integrations with sovereign EU compute reinforces Mistral's full-stack pitch to enterprises and governments rather than competing on raw frontier benchmarks.
  18. 2026-06-25 / VOICESMistral Ships Medium 3.5 (77.6% SWE-Bench Verified), Makes It the Default in Le Chat, and Adds Enterprise ConnectorsOn June 24 Mistral updated Le Chat with Medium 3.5, which it reports scores 77.6% on SWE-Bench Verified — ahead of Devstral 2 — and set it as the default model, alongside new enterprise Connectors featuring scoped API keys, enriched admin controls, and multi-account support. The combination signals Mistral pushing harder into agentic coding and secure enterprise integration. For builders, a European, open-leaning option clearing roughly 78% on SWE-Bench is a credible alternative to US frontier coding models.
  19. 2026-06-24 / NEWSAnthropic Hardens Claude Code With Workload Identity Federation and Sandbox Credential BlockingRecent Claude Code releases (through v2.1.187) add sandbox credential blocking, org-level model restrictions, more reliable structured output and remote MCP, and improved session resume — alongside Anthropic's new Workload Identity Federation (WIF), which replaces static API keys with short-lived, scoped credentials issued per request via existing identities (AWS IAM roles, GCP/Kubernetes service accounts, GitHub Actions tokens, Okta/OIDC). For builders running Claude in automation, WIF removes the 'static key to rotate or leak' problem entirely. The cadence shows how fast Anthropic is hardening Claude Code for production and enterprise use.
  20. 2026-06-24 / NEWSOpenAI Gives ChatGPT Enterprise the Power to Take Actions Inside SlackOn June 23, 2026, OpenAI enabled Slack connector actions for ChatGPT Enterprise and Edu workspaces, letting ChatGPT take real actions in Slack — joining channels, creating reminders, uploading files, and updating a user's profile — not just searching Slack content. The move signals the assistant race shifting from read-only retrieval toward agentic, write-capable actions inside the apps where work already happens. Landing the same day as Anthropic's Claude Tag on Slack, it underscores how fast both labs are racing to own the in-app action layer.
  21. 2026-06-24 / AGENTSOpenAI ships Codex 0.142.0 with token-budget controls and scheduled remindersCodex 0.142.0 (June 22) adds configurable token budgets, usage-limit reset credits, improved plugin browsing, and scheduled UTC reminders. The configurable token-budget control is the most builder-relevant change — it gives finer cost governance over coding-agent runs, a recurring pain point for teams running agents at volume. Incremental but practical for anyone budgeting agent spend.
  22. 2026-06-24 / HACKER NEWSClaude Code 2.1.187 Adds Sandbox Credential Blocking and Org-Enforced Model RestrictionsThe June 24 Claude Code 2.1.187 release introduces a sandbox.credentials setting that prevents sandboxed commands from reading credential files and secret environment variables, plus organization-level model restrictions enforceable through the model picker, CLI flags, and env vars. This is a direct hardening response to the credential-exfiltration and prompt-injection attack class hitting agentic coding tools. The prior 2.1.186 (June 23) added CLI-based MCP login/logout with SSH support and a skills section in the plugin interface.
  23. 2026-06-22 / TOOLSClaude Code v2.1.185 Softens the Stream-Stall Hint — 'Waiting for API response' After 20s, Not 10sThe June 21 release reworks the mid-stream stall notification to read 'Waiting for API response · will retry in …' instead of the alarming 'No response from API · Retrying in …', and only triggers after 20 seconds of silence rather than 10. Small UX change, but it reduces false-alarm anxiety during long reasoning runs where brief silences are normal.
  24. 2026-06-22 / TOOLSCursor 3.7 Adds /in-cloud — Spin Up Cloud Subagents on Dedicated VMs With Reusable SnapshotsCursor 3.7 (June 17) introduced `/in-cloud` to launch cloud subagents on dedicated VMs, with environment setup completing in under 10 minutes and reusable snapshots to speed future runs. Combined with the new Automations layer, it gives solo builders a path to offload long-running or parallel agent work off the laptop without managing infrastructure by hand.
  25. 2026-06-21 / DISPATCHGitHub Releases CC0 Multilingual Repositories Dataset for Non-English Code AIAlongside its June 19 PR-limit launch, GitHub published the Multilingual Repositories Dataset — a CC0 metadata set surfacing public repos with non-English READMEs, issues, and pull requests — to support multilingual AI training, evaluation, and inclusive developer tooling. For builders working on code models or eval suites: a free, permissively-licensed source of non-English code-context data that's otherwise hard to assemble.
  26. 2026-06-21 / DISPATCHGitHub Ships Pull-Request Limits Aimed Squarely at AI-Generated PR SpamOn June 19, GitHub launched configurable pull-request limits letting maintainers cap incoming PRs, allow-list trusted contributors, and specifically throttle AI-generated pull requests. The feature is an explicit response to a flood of low-quality agent-authored PRs overwhelming open-source projects. For builders: expect new friction for bot/agent contributions — get your automation onto bypass lists, and treat unattended PR-generating agents as a maintainer burden, not a freebie.
  27. 2026-06-20 / VOICESClaude Code's June 19 Release Hardens Auto Mode — Destructive git, terraform, and pulumi/cdk Commands Now Blocked Unless You AskClaude Code's June 19 release notes add real guardrails to Auto Mode: it now blocks destructive git commands (reset --hard, checkout -- ., clean -fd, stash drop) when you didn't ask to discard work, blocks git commit --amend on commits the agent didn't make this session, and blocks terraform/pulumi/cdk destroy unless you named the specific stack. For anyone running agents unattended overnight, these directly target the classic 'the agent nuked my uncommitted changes' failure mode. The same release also clears startup clutter and fixes subagent, WebSearch, and background-task bugs.
  28. 2026-06-19 / AGENTSAnthropic adds enterprise-managed MCP connectors with Okta for zero-touch provisioningAnthropic introduced enterprise-managed MCP connector access starting with Okta, letting admins provision a connector once and have users get zero-touch access on first login, with centralized authorization spanning Claude chat, Claude Code, and Cowork for Team and Enterprise plans. This moves MCP from per-user manual setup toward governed, IdP-brokered connector distribution. For builders, it signals MCP maturing into an enterprise-administered control plane rather than an ad-hoc developer integration.
  29. 2026-06-18 / TOOLSTip: Auto-Retry Now Recovers API Drops Mid-Thinking — Stop Babysitting Long Reasoning RunsClaude Code's auto-retry was extended so an API connection drop during the thinking phase automatically retries instead of failing with 'Connection closed while thinking.' Combined with the v2.1.179 mid-stream recovery, long extended-thinking runs and unattended `-p` automations are now meaningfully more resilient to flaky networks — fewer wasted reasoning passes and fewer manual restarts on overnight or CI-driven agent jobs.
  30. 2026-06-17 / TOOLSClaude Code v2.1.179 Ships Mid-Stream Connection Recovery — Partial Responses Now Preserved Instead of Erroring OutFirst seen June 17, v2.1.179 fixes a long-standing failure mode where a dropped streaming connection discarded the entire in-flight response and showed a raw error; partial output is now preserved on reconnect. It also resolves Linux sandbox glob performance, a WSL2/Windows Terminal mouse-wheel scrolling regression, and a feedback survey bug that logged single-digit replies as session ratings. For builders running long autonomous sessions, the streaming-recovery fix directly reduces lost work on flaky networks.
  31. 2026-06-16 / MARKETSSupabase Grafts Agents Onto the BaaS Layer: June Release Adds an AI Coding Agents Plugin, ChatGPT App Integration, and Passkey AuthSupabase's June 2026 release notes add an AI Coding Agents plugin, a ChatGPT app integration, and passkey sign-in — an incumbent backend-as-a-service embedding agent tooling directly rather than ceding the surface to standalone AI app builders. It fits the broader 'graft onto the surface people already use' distribution pattern: instead of users adopting a separate agent IDE, the agent capability ships inside the database/auth platform they're already on. Single-vendor release-note source, so treated as a directional signal rather than a market-shifting move.
  32. 2026-06-15 / SKILLSDisable the 'workflow' keyword trigger in Claude Code /config so the word stops auto-launching dynamic workflowsA June 2026 Claude Code update added a 'Workflow keyword trigger' setting in /config that lets you stop the literal word 'workflow' in a prompt from kicking off a dynamic multi-agent workflow. If you discuss workflows conversationally (or your CLAUDE.md mentions them), this prevents accidental, token-heavy orchestration runs. Builder move: turn it off in repos where you talk about workflows but only want orchestration on explicit command.
  33. 2026-06-14 / AGENTSClaude Code June changelog: plugin management, version pinning, and richer subagent feedbackAnthropic's June Claude Code updates add a /plugin list command (with --enabled/--disabled filters), requiredMinimumVersion/requiredMaximumVersion managed settings that refuse to start outside an allowed range, and Stop/SubagentStop hooks that can return hookSpecificOutput.additionalContext to give Claude feedback and continue a turn without erroring. Skills also gained an escape syntax for a literal $ before a digit. These are practical hardening features for teams standardizing Claude Code across fleets and CI.
  34. 2026-06-14 / SOURCESOpenAI Ships ChatGPT 'Library' for Enterprise/Edu/Healthcare Plus New Cloud Console Admin ControlsOpenAI's June 12 release adds a shared 'Library' for ChatGPT Enterprise, Edu, and Healthcare workspaces plus new admin controls in the Cloud Console. It's an enterprise-governance play — managed, shared asset libraries — rather than a model-capability change. Relevant if you're standardizing ChatGPT deployment and access policy across a team.
  35. 2026-06-14 / TOOLSCursor Ships Composer 2.5 and a Premium Seat Tier for Heavy Agent UsersCursor's June update introduces Composer 2.5 — its latest in-house model claiming frontier performance at a fraction of the cost — plus a new Premium seat aimed at developers running many agents, with improved spend forecasting and controls. Enterprise admins get organization-level controls to manage multiple teams from one place with separate security, governance, and budget settings, and new Groups for flexible access permissions. This continues Cursor's push toward fleet-scale agent management (up to 8 parallel agents via Mission Control).
  36. 2026-06-12 / SKILLSPin extended-thinking cost to zero on routine turns with MAX_THINKING_TOKENS=0Claude Code v2.1.166 made MAX_THINKING_TOKENS=0 and --thinking disabled work reliably, so you can switch off extended reasoning on mechanical or low-stakes turns where it only adds latency and token spend. Reserve thinking budget for genuinely hard planning and debugging steps and disable it for boilerplate edits, formatting, and scripted runs. For unattended scheduled agents this is a direct, predictable lever on per-run cost.
  37. 2026-06-12 / SKILLSUse Dynamic Workflows to orchestrate one big task across many background sub-agentsClaude Code's Dynamic Workflows (research preview, Week 22) let you script deterministic fan-out — loops, conditionals, and pipelines that dispatch dozens of background sub-agents and synthesize their structured outputs — rather than hoping one model self-organizes a large task. It's the right tool when you need comprehensiveness (decompose and cover in parallel) or confidence (independent perspectives plus an adversarial verify pass) at a scale one context window can't hold. Encode the structure in the script: what fans out, what verifies, what merges.
  38. 2026-06-12 / SKILLSRun long shell jobs as attachable background tasks with the `!` prefix in agent viewClaude Code's agent view lets you prefix a shell command with `!` to launch it as a background job you can detach from and re-attach later, instead of blocking the session on a long build, test suite, or dev server. The same view dispatches and monitors background agent sessions from one terminal dashboard, so you start work, send it to the background, and only jump back in when input is needed. This is the practical primitive for running several parallel agent sessions without juggling terminal tabs.
  39. 2026-06-12 / SKILLSHide built-in commands from the model with disableBundledSkills to reclaim contextClaude Code v2.1.169 added a disableBundledSkills setting that stops the bundled skills/commands from being advertised to the model, trimming the always-on token overhead when your project only uses custom skills. Combined with the new /reload-skills and /reload-plugins commands, you can iterate on extensions mid-session without restarting and losing prompt cache. This is a direct lever on the 'distraction' failure mode — fewer irrelevant tool descriptions competing for the model's attention.
  40. 2026-06-12 / SKILLSConfigure a fallbackModel chain so a rejected turn retries instead of dyingClaude Code v2.1.166 added a fallbackModel setting that takes up to three models tried in order, and the harness now auto-retries a turn once on the fallback when the API throws an unexpected non-retryable error (auth, rate-limit, request-size, and transport errors still surface immediately). For long unattended runs and scheduled agents this turns a single transient model hiccup into a silent recovery instead of a dead session. Pair it with cheaper models lower in the chain to bound cost when the primary is overloaded.
Open latest cited source