The daily firehose — everything the toolchain shipped today, already filtered.
We read every release from the 1309 tools on our watchlist at the source — GitHub and GitLab release notes, vendor release pages and changelogs, project blogs and feeds, vendor press releases, and the source code behind the tag. Bug-fix-only releases and non-product newsroom noise are dropped; what's left is summarized down to the new capability, how to try it, and any screenshots or videos the release itself published. Every entry links to the sources it was built from.
READ AS
Focus:What features shipped and what problems they solve, plus launches, packaging and where the market is moving.
Leaves out:Command-line flags, API-only deltas, usage examples, and releases that are only refinements.
Focus:New security controls, new APIs, and product improvements you can put to work, plus anything an upgrade breaks.
Leaves out:Launch and marketing news, governance, pricing, and documentation changes with no operational effect.
Focus:Dashboards, reporting, and anything you can quantify: metrics, coverage, risk scores, compliance impact, vendor moves.
Leaves out:Settings, flags, API and command-line detail, library bumps, and minor releases with no reporting or posture angle.
Focus:API, SDK and MCP changes, new libraries, breaking changes in what you import, and cyber tooling.
Leaves out:Governance, SOC operations, and vendor launches with no developer surface.
AMOUNT
DETAIL
ORDER
This issue came out before editions were ranked, so it shows the Standard edition in issue order.
BloodHound v9.4.0 adds ADCSADCSActive Directory Certificate Services, Microsoft's PKI role for issuing and managing X.509 certificates inside a Windows domain; cyber tools target it because misconfigurations frequently allow privilege escalation or domain compromise. ESC14 detection, new graph/relationship APIs, webhook creation, auditor role expansions, and ingest metrics.
└──▷ GET THIS VERSION
$ git clone --branch v9.4.0 https://github.com/SpecterOps/BloodHound.git
# already have the repo? check out this version:$ git checkout v9.4.0
Adds ADCS ESC14 Scenario A attack pathattack pathA possible sequence of steps an attacker could take to reach a valuable target. It connects weaknesses, permissions and reachable systems rather than treating each finding separately. coverage for certificate-based privilege escalationprivilege escalationGaining permissions beyond those an account or process initially had. It can turn limited access into administrator control or access to another user's resources. detection.
New GET api/v2/relationships/{relationship_id} endpoint for retrieving individual graph relationship details.
New GET node endpoint (api/v2/...) for retrieving individual graph node details.
Introduces webhook creation capability for event-driven alerting integrations.
Adds data_quality_stats table and supporting APIs for tracking data quality statistics.
+12 moreshow less
Adds data_quality_aggregations table migration to support aggregated data quality reporting.
Expands Auditor role permissions to include SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. configuration page access.
Expands Auditor role permissions to include file ingest visibility.
Adds explicit permissions for managing and reading ingest files.
New queue latency metric for ingest tasks, enabling observability into ingest pipeline delays.
New ingest task counter metric for tracking volume of ingest operations.
Alerts page scaffolding introduced as a new UI surface for future alerting workflows.
Disallows creation of new custom Cypher rules when the query yields no results, preventing misconfigured detections.
Adds a support request bundle button for easier diagnostics submission.
Introduces Variable Analysis Mode pipeline steps for more flexible analysis configurations.
Adds collector file service for downloading collector binaries directly from the platform.
Adds screen reader and accessibility (a11y) support across graph controls and admin pages.
A humble and fast security-oriented HTTP headers analyzer.
humble 1.63 adds HARHARAn HTTP Archive file, a JSON-format log of all network requests and responses captured by a browser's developer tools, used by cyber tools to replay, inspect, or analyze web traffic without re-running a live session. file analysis via -if and a new insecure header check.
└──▷ GET THIS VERSION
$ git clone --branch 1.63 https://github.com/rfc-st/humble.git
# already have the repo? check out this version:$ git checkout 1.63
└──▷ TRY IT
Analyze a captured HAR file from a browser session to audit its HTTP response headers without making live requests.
$ python humble.py -if <path/to/capture.har>
Supports analyzing HAR (HTTP Archive) files as input via the -if option, enabling offline inspection of captured browser traffic.
Adds 1 new insecure header check, expanding detection coverage.
Wiz Defend is a cloud detection and response platform that identifies and helps remediate real-time threats across cloud environments.
Wiz Defend adds AI coding assistant context-aware detections for symlink and SSH key attacks at clone and write time.
Detects malicious symlinks pointing to sensitive files at git clone time, catching GhostApproval-style attacks before AI coding assistant involvement.
Monitors writes to sensitive files (e.g., SSH authorized_keys) at write time, triggering alerts regardless of whether the write originates via a symlink.
Adds AI agent context awareness: alerts auto-elevate severity when malicious activity originates from an AI coding assistant process operating on untrusted repositories.
Leverages Wiz Runtime Sensor telemetry to identify active endpoints and their exposure based on inbound network traffic.
Wiz Cloud is a cloud security platform that scans and visualizes risks across cloud environments, helping teams identify vulnerabilities, misconfigurations, and threats.
Wiz ASMASMAttack Surface Management: the continuous practice of discovering, inventorying, and monitoring an organization's externally exposed assets so that cyber tools can flag unknown or misconfigured entry points before attackers find them. adds automated subdomain recon, API endpoint discovery, runtime sensor telemetry, and custom target input to map attack surfaceattack surfaceThe exposed systems, interfaces and access points an attacker could try to use. It includes internet-facing services as well as identities, software and other reachable resources..
└──▷ HOW TO FIND IT
Extend ASM coverage to legacy or air-gapped systems not reached by automated discovery by manually specifying their addresses.
📍In the Wiz console, go to ASM › Custom Targets and add the IP addresses, domains, or IP ranges of legacy or isolated systems you want analyzed.
Adds automated subdomain reconnaissance seeded from owned root domains to surface shadow exposures.
Discovers REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. endpoints via cloud-native API gateways, Wiz Runtime Sensor telemetry, and Attack Surface Scanner-detected API specifications.
Leverages Wiz Runtime Sensor inbound network traffic telemetry to identify and assess active endpoint exposure.
Supports manual input of IP addresses, domains, and IP ranges as Custom Targets to cover legacy or isolated systems outside automated discovery.
Maps cloud network configurations to calculate effective exposure for every cloud object.
Zscaler Posture Control is a cloud-native application protection platform (CNAPP) that identifies and remediates security risks, misconfigurations, and compliance violations across multi-cloud environments.
Zscaler Posture Control adds AI-SPMAI-SPMAI Security Posture Management: a discipline for continuously discovering, assessing, and remediating risks across AI pipelines, models, and data, giving cyber tools a structured way to govern AI infrastructure the way CSPM governs cloud. with shadow AI discovery, drift detection, and automated red teaming mapped to MITRE ATLASMITRE ATLASA MITRE-maintained knowledge base of adversarial tactics and techniques targeting AI and machine learning systems, modelled on ATT&CK, giving cyber tools a structured taxonomy for AI-specific threat modelling., NIST AI RMFNIST AI RMFA voluntary framework published by NIST that gives organizations a structured process for identifying, assessing, and managing risks introduced by AI systems, giving cyber tools a common vocabulary and governance scaffold., and OWASP LLM Top 10OWASP LLM Top 10A ranked list of the ten most critical security risks specific to large language model applications, maintained by OWASP, giving cyber tool developers a shared taxonomy for prioritizing LLM attack surfaces..
New AI Security Posture Management (AI-SPM) provides continuous discovery and posture assessment of AI assets including shadow AI, unsanctioned models, embedded SaaSSaaSSoftware delivered over the internet and run on the provider's infrastructure, so the buyer pays a subscription instead of installing or operating anything locally. AI, developer-built agents, and MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers.
New AI Bill of MaterialsBill of MaterialsA structured list of every component, library, and dependency that makes up a piece of software or hardware, used by cyber tools to track what is inside a product and surface known vulnerabilities in its parts. (AI-BOMAI-BOMAn inventory of every AI model, dataset, and dependency used in a system, analogous to a software bill of materials, giving cyber tools a structured way to audit AI supply-chain risk.) maps model lineage across discovered AI assets for full supply-chain visibility.
Adds posture assessment for AI-specific risks: misconfigurations, excessive permissions, vulnerable RAGRAGRetrieval-augmented generation: looking up relevant documents and adding them to the prompt before the model answers. It grounds responses in private or current data without retraining. frameworks, and exposed data pipelines.
New AI posture drift detection continuously compares permission scope, configuration state, and data access patterns against established baselines to surface slow-erosion changes point-in-time assessments miss.
Automated red teaming runs continuous, high-scale tests across the AI lifecycle and maps findings to MITREMITREA US federally funded research and development center that maintains public frameworks—ATT&CK, CVE, D3FEND—that cyber tools use to classify threats, vulnerabilities, and defensive techniques in a common language. ATLAS, NIST AI RMF, OWASPOWASPA nonprofit foundation that publishes open standards, guides, and ranked vulnerability lists—most famously the OWASP Top 10—giving cyber tools a vendor-neutral benchmark to test or report against.LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code. Top 10, and the EU AI ActEU AI ActEU regulation that classifies AI systems by risk level and imposes compliance obligations—documentation, transparency, human oversight—on developers and deployers, shaping how AI-based cyber tools are built and sold in Europe..
+1 moreshow less
Remediation workflows integrate with ITSMITSMIT Service Management: a framework of practices for delivering and supporting IT services, covering ticketing, change control, and incident response. Cyber tools integrate with it to automate alert-to-ticket workflows. (Jira, ServiceNow), DLPDLPData Loss Prevention: a category of controls that inspect and block sensitive data from leaving an organisation via email, web, or removable media, giving cyber tools a policy enforcement layer over data in motion and at rest., and DSPMDSPMData security posture management, locating sensitive data and checking how it is protected and who can reach it. It helps teams prioritize risks around their data stores. platforms, with prioritized alerts and least-privilege access enforcement.
Palo Alto Prisma Cloud is a cloud-native security platform that protects applications, data, and infrastructure across multi-cloud environments throughout the development lifecycle.
Prisma AIRS 3.0 adds Japan local cloud region, AI Agent SSPMSSPMSaaS security posture management, checking hosted applications for risky settings, excessive access and policy violations. It helps teams manage risks across their subscribed applications., AI Red Teaming, and expanded model/runtime scanning.
New local cloud region in Japan for Prisma AIRS, enabling data residency and low-latency processing for Japanese enterprises.
New AI Agent SSPM (SaaSSaaSSoftware delivered over the internet and run on the provider's infrastructure, so the buyer pays a subscription instead of installing or operating anything locally. Security Posture Management) detects and governs agentic threats including identity impersonation, memory manipulation, and tool misuse across autonomous AI assistants.
New AI Red Teaming capability performs automated, adaptive penetration tests against AI deployments using an agent that learns and evolves like a real attacker.
New AI Model Security scanning checks third-party models for vulnerabilities, malicious scripts, tampering, and deserializationdeserializationTurning stored or transmitted data back into objects a program can use. Unsafe handling of untrusted data can allow unexpected behavior or code execution. attacks before deployment.
New AI Runtime Security protection for LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code.-powered apps covers prompt injectionprompt injectionAn attack where text the model reads, such as a web page, file or tool result, carries instructions that override the user's. It is the main security risk for agents with tools., sensitive data leaks, and Indirect Prompt Injection (IDPI) embedded in websites.
+1 moreshow less
New Prisma AIRS AI Gateway centralizes runtime security, data protection, and agent identity controls across all AI tools in an organization from a single control plane.
sigma-cli v3.1.0 adds JUnit XMLJUnit XMLA widely adopted test-report file format, originally from the Java JUnit framework, that records test suite results as XML; cyber tools emit it so their output plugs into CI systems like Jenkins or GitLab without custom parsers. export and per-rule file output with directory structure preservation.
└──▷ GET THIS VERSION
$ git clone --branch v3.1.0 https://github.com/SigmaHQ/sigma-cli.git
# already have the repo? check out this version:$ git checkout v3.1.0
└──▷ TRY IT
Output each converted rule as its own file mirroring the source directory layout — useful for version-controlled rule repos or selective SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. deployment.
Adds JUnit XML export for sigma convert results, enabling CI pipeline integration with standard test reporting tools.
Adds --output-dir and --output-filename-template flags to write each converted rule to a separate file, with optional preservation of the source directory hierarchy.
Qualys Patch Management is a cloud-based tool that automates the discovery, prioritization, and deployment of patches across IT assets to reduce vulnerabilities.
Qualys Patch Management adds autonomous zero-dayzero-dayA vulnerability for which defenders have had no prior opportunity to apply a fix, often because it was previously unknown. A zero-day exploit takes advantage of that gap. remediation via TruRisk Eliminate and Windows 11 upgrade orchestration through QQL and CAR scripts.
└──▷ HOW TO FIND IT
Identify and deploy Windows 11 enablement packages at scale to upgrade endpoints already running recent Windows 11 builds with minimal operational impact.
📍In TruRisk Eliminate Patch Catalog, run QQL: patch.title:"Enablement" — select the matching patch, create a patch job, and target the relevant asset groups.
Find ISOISOA widely adopted international standards body that publishes technical specifications; cyber tools cite ISO standards to signal compliance with recognized benchmarks for quality, security, or interoperability.-based Windows feature update patches for environments where enablement packages are unavailable (e.g., 23H2 → 24H2 transition).
📍In TruRisk Eliminate Patch Catalog, run QQL: patch.isFeatureUpdate:true AND patch.vendor:Microsoft — enable the required ISO package, then create and deploy a patch job to the target asset groups.
Directly upgrade Windows 10 endpoints to the latest Windows 11 build without intermediate steps using a pre-built CAR library script.
📍In TruRisk Eliminate, navigate to Script Library → System Updates and Upgrades → Windows OS Upgrades, select the Windows 10 to Windows 11 upgrade script, import it in approved mode, and execute it on the required asset groups.
New autonomous remediation workflows via Qualys TruRisk Eliminate enable machine-speed patching, mitigation, container image replacement, workload isolation, and firewall policy updates triggered directly from confirmed vulnerability findings.
Supports patchless/compensating mitigation actions when immediate patching is not feasible, reducing exposure window while permanent fixes are prepared.
Enables enterprise-wide Windows 11 upgrade orchestration through TruRisk Eliminate, covering three paths: enablement packages, ISO-based feature updates, and direct Windows 10-to-Windows 11 upgrades.
New QQL filters for patch targeting: patch.title:"Enablement" for enablement packages and patch.isFeatureUpdate:true AND patch.vendor:\Microsoft\ for ISO-based feature updates.
Windows 10-to-Windows 11 upgrade scripts now available in the Custom Assessment and Remediation (CAR) library under Script Library → System Updates and Upgrades → Windows OS Upgrades.
Qualys CyberSecurity Asset Management provides continuous visibility into all IT assets, identifying security gaps, risks, and unauthorized devices across hybrid environments.
Qualys CSAM integrates with Cisco Cloud Control, exposing asset inventory, TruRisk scores, and remediation workflows to Cisco AI agents.
└──▷ HOW TO FIND IT
Use Cisco AI Canvas to query Qualys asset and risk context for a specific host directly from the AI agent workflow.
📍In Cisco AI Canvas, navigate to the AI Agents panel, select the Qualys integration tools (Unified Asset Inventory, TruRisk Prioritized Findings), and invoke a tool call for a target host to retrieve vulnerability detections, TruRisk score, exploit context, and available remediation workflows.
New Cisco Cloud Control integration connects Qualys Unified Asset Inventory, TruRisk Prioritized Findings, and remediation workflows into Cisco's AI Canvas environment.
Cisco AI agents can now query Qualys at runtime — pulling asset inventory, vulnerability detections per host/workload, TruRisk scores, exploit context, and triggering remediation or mitigation workflows as typed tool calls.
Enables cross-domain asset and risk visibility spanning cloud, network, identity, datacenter, and security domains without manual tool-switching.
Zscaler Data Protection secures sensitive data across cloud applications, endpoints, and networks by enforcing DLP policies and preventing unauthorized data exfiltration.
Zscaler Workflow Automation adds end-user DLPDLPData Loss Prevention: a category of controls that inspect and block sensitive data from leaving an organisation via email, web, or removable media, giving cyber tools a policy enforcement layer over data in motion and at rest. triage, self-service email quarantine release, and automated exception management.
└──▷ HOW TO FIND IT
Allow end users to self-release quarantined emails after justification, eliminating helpdesk tickets for routine outbound email DLP hits.
📍In the Zscaler console, go to Advanced Account Settings and enable Enable Email Quarantine Release for End Users.
Manually release a quarantined email to specific recipients from the incident details view when admin review is required.
📍In the Zscaler console, open the incident in the advanced incident details interface, then select the Release Email Quarantine action and choose specific recipients to deliver the message to.
Adds end-user justification workflows that engage users directly via Slack, Microsoft Teams, or email when an inline DLP policy triggers, routing responses to auto-remediation, manager escalation, or analyst enrichment.
Enables manager-approved, time-bounded DLP policy exceptions created automatically through workflow responses — no manual policy edits required.
Introduces Release Email Quarantine action in the advanced incident details interface, letting admins deliver quarantined messages to all or specific recipients directly from the platform.
New Enable Email Quarantine Release for End Users setting in Advanced Account Settings allows end users to release their own quarantined emails after providing a business justification or obtaining manager approval.
Supports native integration with ServiceNow and Jira to auto-create enriched incident tickets with user justification and forensic context attached.
+2 moreshow less
Supports customizable, cloneable, and translatable end-user justification survey templates for global workforce coverage.
Integrates identity-provider user-to-manager mapping to enable automated manager escalation routing within DLP incident workflows.
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
Claude Code v2.1.205 upgrades /doctor into a full setup checkup, hardens auto mode safety rules, and improves the agents view.
└──▷ GET THIS VERSION
$ git clone --branch v2.1.205 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:$ git checkout v2.1.205
└──▷ TRY IT
Run a full environment health check and let Claude Code automatically fix any issues it finds.
$ /doctor
Review the enriched agent list — with colored state labels, classifier headlines, and PR links — for all background sessions.
$ claude agents
Adds auto mode rule that blocks tampering with session transcript files.
Improves auto mode to prompt for confirmation before running rm -rf on an unresolvable variable, reducing accidental data loss.
Upgrades /doctor into a full setup checkup that can both diagnose and fix issues; adds /checkup as an alias.
Enhances claude agents view: PR-linked sessions (edit, merge, comment, push) now surface the PR link directly in the list.
Enhances claude agents view: rows show a colored state word and classifier-written headline instead of raw tool call text; peek view shows full status and the exact ask for blocked sessions.
+3 moreshow less
Background task notifications now explicitly state no human input has occurred, preventing fabricated in-transcript approvals from being acted on.
Streams auto-update binary downloads to disk instead of buffering in memory, cutting peak updater memory usage by ~400 MB.
Reserves the Claude BrowserMCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server name (alongside Claude Preview); user-configured MCP servers can no longer register under either name.
8 details in Everything
└──▷ BREAKING ON UPGRADE
!User-configured MCP servers can no longer register under the names Claude Browser or Claude Preview
1 more release in this issue
· 2026-07-07 → 2026-07-08
Claude Code v2.1.203 adds login-expiry warnings, manual-mode visibility, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. roots updates, and a VSCode remote-control toggle.
└──▷ GET THIS VERSION
$ git clone --branch v2.1.203 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:$ git checkout v2.1.203
Adds a pre-expiry warning when your login session is about to expire, letting you re-authenticate before background sessions are interrupted.
Adds a grey ⏸ badge in the footer when manual permission mode is active, so the current permission mode is always visible.
Exposes the session's additional working directories in MCP roots/list, and sends notifications/roots/list_changed whenever that set changes — enabling MCP servers to stay in sync with multi-root workspaces.
Changes the Escape key (instead of left arrow) to close background tasks, diff, and workflow detail views — practitioners must update muscle memory.
Changes the empty claude agents view to always display organized sections (Needs Input / Working / Completed) with descriptions, even when no agents are running.
+2 moreshow less
Moves startup 'claude command missing or broken' warnings out of the startup flow into /doctor and /status only.
[VSCode] Adds a Settings toggle for Enable Remote Control for all sessions.
7 details in Everything
└──▷ BREAKING ON UPGRADE
!The left arrow key no longer closes the background tasks, diff, and workflow detail views — press Esc instead
Lightweight coding agent that runs in your terminal
Codex CLI v0.143.0 adds remote plugins by default, system proxy support, daemon pairing codes, new Bedrock models, and richer app-server APIs.
└──▷ GET THIS VERSION
$ git clone --branch rust-v0.143.0 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.143.0
└──▷ TRY IT
Generate a manual pairing code from a running Codex daemon to connect a remote-control client without automatic discovery.
$ codex remote-control pair
Enables remote plugins by default, with richer catalog rows, npm marketplace sources, and visible remote/local version info.
Supports routing authentication and Responses API traffic through macOS and Windows system proxies, including PACPACA Proxy Auto-Configuration file, a JavaScript snippet a browser or OS fetches to decide which proxy server to route each request through, letting cyber tools intercept or redirect traffic without per-app configuration. and WPAD configurations.
Adds codex remote-control pair subcommand for generating manual pairing codes from a running daemon.
Adds Amazon Bedrock GPT-5.6 Sol, Terra, and Luna models with first-class support for max reasoning effort.
Enables MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. tool search by default; ChatGPT-hosted MCP servers can now explicitly use session authentication.
+1 moreshow less
Expands app-server client API to inspect environments, list descendant threads by ancestor, and fork history through a specific turn.
6 details in Everything
0 more releases in this issue3 more releases in this issue · 2026-07-07 → 2026-07-08
Adds Ultra concurrency warning, proxy-aware WebSocketWebSocketA protocol that upgrades an HTTP connection into a persistent, full-duplex channel, letting a server push data to a client without repeated polling. Cyber tools use it for real-time event streaming and live command output. connector, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. tool snapshot reuse, and Windows sandbox runtime access.
└──▷ GET THIS VERSION
$ git clone --branch rust-v0.144.0-alpha.2 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.144.0-alpha.2
Adds a TUITUIA text-based user interface that runs inside a terminal, using characters and ANSI escape codes to render menus, panes, and widgets without a graphical desktop, letting cyber tools ship rich interactivity with no GUI dependency. warning when Ultra reasoning is selected with high multi-agent concurrency (≥8 threads), displaying configured thread and subagent limits so users understand cost implications.
Adds a proxy-aware WebSocket connector that respects outbound proxy policy and custom-CA trust, matching the same effective HTTP proxy configuration used for REST calls.
Reuses a single MCP tool snapshot within each sampling request, reducing latency and ensuring consistent tool state between Apps World State and tool-router construction.
Grants Windows sandbox sessions read/execute access to the managed primary runtime cache (%USERPROFILE%\.cache\codex-runtimes), enabling bundled Python, Node, and native tools in elevated sandbox contexts.
$ git clone --branch rust-v0.144.0-alpha.1 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.144.0-alpha.1
└──▷ TRY IT
Verify that a pnpm-managed Codex install is recognized correctly in your environment.
$ codex doctor
Detects Codex installs managed by pnpm, surfacing pnpm status in codex doctor output alongside npm and bun.
Speeds up the /review branch picker by switching branch enumeration to git for-each-ref scoped to refs/heads, preventing timeouts in large repos.
Adds trace spans around configured hook command execution, making slow installed hooks visible directly in traces.
Supports externally provided Codex auth snapshots with explicit runtime capabilities, installable via the existing ExternalAuth provider path.
Adds writes approval mode, serialized MCP OAuthMCP OAuthAn authentication layer for the Model Context Protocol that lets MCP clients obtain delegated access tokens so cyber tools can call MCP servers on a user's behalf without handling raw credentials. credential stores, canonical turn item emission, and improved Windows URL safety detection.
└──▷ GET THIS VERSION
$ git clone --branch rust-v0.143.0-alpha.39 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.143.0-alpha.39
└──▷ TRY IT
Set the new writes approval mode on an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. tool so the agent auto-approves read operations but prompts only for writes.
$ codex mcp get
Adds writes approval mode for MCP app tool approval, giving practitioners a middle ground between auto and prompt that gates only write operations
Emits canonical TurnItem lifecycle events for command execution, dynamic tool calls, sub-agent activity, collab tool calls, and collab wait items, enabling more structured agent event streams
Enables auth elicitation by default, so authentication prompts are surfaced automatically without manual configuration
Supports hosted success redirects in the login flow
Shows reset credit details (title, expiry, idempotency key) in the TUITUIA text-based user interface that runs inside a terminal, using characters and ANSI escape codes to render menus, panes, and widgets without a graphical desktop, letting cyber tools ship rich interactivity with no GUI dependency. redemption picker so users can choose exactly which credit to consume
+2 moreshow less
Extends Windows dangerous-command detection to handle mixed-case URLs (e.g., HTTP:// or hTtPs://) in PowerShell commands
Adds connector runtime latency metrics for observability into MCP connector performance
OpenHands cloud-1.43.0 adds SMTPSMTPThe application-layer protocol used to send email between servers and from clients to servers; cyber tools use it to relay alerts, phishing simulations, or exfiltrated data over standard mail infrastructure. email service, user login tracking, and a renamed Usage & Monitoring dashboard.
└──▷ GET THIS VERSION
$ git clone --branch cloud-1.43.0 https://github.com/OpenHands/OpenHands.git
# already have the repo? check out this version:$ git checkout cloud-1.43.0
└──▷ HOW TO FIND IT
Check whether email delivery (SMTP or Resend) is enabled for your OpenHands Cloud instance — useful before configuring notification workflows.
📍In the console, go to Usage & Monitoring and look for the email-enabled status indicator showing SMTP or Resend as active.
Adds SMTP email service integration, enabling transactional email delivery alongside the existing Resend provider.
New email-enabled status surface shows whether SMTP or Resend email is active in the platform configuration.
Tracks user login timestamps, giving admins visibility into last-active data per user.
Renames the Admin Dashboard to Usage & Monitoring for clearer intent.
4 details in Everything
3 more releases in this issue
· 2026-07-07 → 2026-07-08
OpenHands 1.10.0 adds SMTPSMTPThe application-layer protocol used to send email between servers and from clients to servers; cyber tools use it to relay alerts, phishing simulations, or exfiltrated data over standard mail infrastructure. email service, user login tracking, and a renamed Usage & Monitoring dashboard.
└──▷ GET THIS VERSION
$ git clone --branch 1.10.0 https://github.com/OpenHands/OpenHands.git
# already have the repo? check out this version:$ git checkout 1.10.0
└──▷ HOW TO FIND IT
Check whether email delivery is active for your deployment — useful when onboarding users who need invite or notification emails.
📍In the console, go to Settings › Usage & Monitoring and check the Email Enabled indicator for your configured SMTP or Resend provider.
Adds SMTP email service integration, enabling email delivery alongside the existing Resend provider.
Surfaces email-enabled status in the UI for both SMTP and Resend configurations.
Tracks user login timestamps, giving admins visibility into account activity.
Renames the admin dashboard to Usage & Monitoring for clearer intent.
OpenHands cloud-1.42.0 adds a super-admin management API for granting, revoking, and listing admin privileges.
└──▷ GET THIS VERSION
$ git clone --branch cloud-1.42.0 https://github.com/OpenHands/OpenHands.git
# already have the repo? check out this version:$ git checkout cloud-1.42.0
New enterprise super-admin management endpoint enables granting, revoking, and listing super-admin roles via API.
An open-source AI agent that brings the power of Gemini directly into your terminal.
Gemini CLI v0.50.0 adds tool registry discovery for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. integrations.
└──▷ GET THIS VERSION
$ git clone --branch v0.50.0 https://github.com/google-gemini/gemini-cli.git
# already have the repo? check out this version:$ git checkout v0.50.0
Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.
CrewAI 1.15.2 adds inline skill definitions, templated Flow action inputs, stream frame protocol, and AgentExecutor feedback handling.
└──▷ GET THIS VERSION
$ git clone --branch 1.15.2 https://github.com/crewAIInc/crewAI.git
# already have the repo? check out this version:$ git checkout 1.15.2
Supports inline skill definitions, eliminating the need for separate skill files in crew/flow configurations.
Adds templated Flow action inputs, enabling dynamic parameterization of Flow actions at runtime.
Defines a stream frame protocol for Flows, enabling structured streaming output from Flow executions.
Adds a generated Flow Definition authoring skill to guide AI coding agents when scaffolding Flows.
Implements message setup and feedback handling in AgentExecutor for richer agent interaction loops.
+4 moreshow less
Adds repository agents to Flow definitions, enabling agent-backed steps directly in Flow authoring.
Adds text helpers for Flow CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. prompts and skill examples, streamlining expression authoring in Flows.
Types tool and app fields in CrewDefinition, providing stronger schema validation for crew configs.
Pulls latest LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code. models dynamically in the crew wizard, keeping model selection current without manual updates.
Get up and running with Kimi-K2.6, GLM-5.1, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Ollama v0.31.2 adds flash attention on Pascal GPUsGPUsA graphics processing unit, a processor that performs many calculations in parallel. AI workloads use GPUs to speed up model training and inference., iGPU vision-model offload, and privacy-hardened Claude Code launch.
└──▷ GET THIS VERSION
$ git clone --branch v0.31.2 https://github.com/ollama/ollama.git
# already have the repo? check out this version:$ git checkout v0.31.2
└──▷ TRY IT
Launch Claude Code backed by a local model with telemetry and feedback reporting disabled by default (no extra flags needed — privacy-off is now the default).
$ ollama launch claude
Enables flash attention on older NVIDIA Pascal GPUs (compute capability 6.x), unlocking faster inference on GTX 10-series and similar cards.
Enables iGPU to offload vision models with padding to fit available memory, broadening multimodalmultimodalAble to work with more than one type of input or output, such as text, images and audio. A multimodal model can combine information across those formats. support on integrated graphics.
Disables Claude Code telemetry, error reporting, feedback commands, and non-essential traffic by default when using ollama launch claude.
Updates the MLXMLXApple's machine-learning framework for Apple silicon, which uses the unified memory shared by CPU and GPU to run and fine-tune models on a Mac. and llama.cpp engines to their latest upstream versions.
THE CYBER TOOLCHAIN