PAGES
Subscribe Get it delivered — five to ten releases we picked each morning from the tools you run, with every release in the issue on the web. Compare plans →

The Cyber Toolchain — issue 012, July 11, 2026

THE CYBER TOOLCHAIN NO. 012
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED JULY 11, 2026 · EVERY DAY
EDITIONS tail grep head ls diff uniq

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.

VIEW
ISSUE VIEW scroll
ISSUE WIDTH reading width
Do you prefer scroll?
ADDITIONAL FILTERS
$ tct list   # 15 tools matched
Reverse Engineering, Malware & Forensics
◆  Reverse Engineering

jadx

Sources Release notes → v1.5.6 Notes

Dex to Java decompiler

jadx v1.5.6 adds graph views with code pane sync, a call graph library, and resource tree improvements.

└──▷ GET THIS VERSION
$ git clone --branch v1.5.6 https://github.com/skylot/jadx.git
# already have the repo? check out this version:
$ git checkout v1.5.6
  • Adds graph views with code pane synchronization in the GUIGUIA graphical user interface, controls such as windows, buttons and menus. It lets people use software visually rather than entering commands. for visual code flow analysis.
  • New call graph module/library enables programmatic traversal of method call relationships.
  • Collapses single-child directory chains in the resource tree for cleaner navigation.
  • Adds Copy Reference option to the context menu for quick symbol referencing.
  • Supports pasting file lists directly into the file chooser dialog.
+1 more show less
  • Opens the single loaded class automatically after load, reducing manual navigation steps.

6 details in Everything

Was this useful?
Network Security
◆  Firewalls & Network Defense

Cloudflare

Sources Blog / feed → Improving Smart Tiered Cache for Public Cloud Regions (2026-07-10) Blog

Cloudflare Blog is an official publication sharing updates, technical insights, and announcements about Cloudflare's internet security, performance, and networking products and services.

Cloudflare Smart Tiered Cache now supports cloud region hints for AWS, GCP, Azure, and Oracle Cloud anycast origins.

└──▷ HOW TO FIND IT
Set a region hint for an anycast AWS origin so Smart Tiered Cache routes through the optimal upper tier instead of hairpinning across continents.
📍In the Cloudflare dashboard, go to Caching › Tiered Cache › Origin Configuration. Find your anycast origin IP, click the edit icon, select Set Region Hint, and choose your region (e.g., aws:us-east-1). Save to apply.
  • Adds cloud region hint configuration for anycast origins, letting Smart Tiered Cache select optimal upper tiers for origins on AWS, GCP, Azure, and Oracle Cloud.
  • New Origin Configuration table in the dashboard (Caching > Tiered Cache > Origin Configuration) enables per-IP region hint assignment or bulk editing of all origin IPs at once.
  • Supports region hint management via Cloudflare API and Terraform for infrastructure-as-code workflows.
  • Introduces anycast origin detection using speed-of-light latency checks across multiple global checkpoint data centers.
  • Automatically refreshes cloud provider IP range mappings every few hours and upper-tier probe data every 15 minutes, with geographic fallback for regions lacking probe data.

5 details in Everything

Was this useful?
Cloud & Container Security
◆  Container & Image Security

Docker Compose

Sources Release notes → v5.3.0 2 releases · 2026-06-23 → 2026-07-02 Notes LatestStable

Define and run multi-container applications with Docker

Docker Compose v5.3.0 adds native support for pre-start init containers.

└──▷ GET THIS VERSION
$ git clone --branch v5.3.0 https://github.com/docker/compose.git
# already have the repo? check out this version:
$ git checkout v5.3.0
└──▷ USE IT
Run a setup task (e.g., DB migration) in an init container that must succeed before your app service starts.
yaml
services:
  app:
    image: myapp:latest
    depends_on:
      init-migrate:
        condition: service_completed_successfully
  init-migrate:
    image: myapp:latest
    command: ["python", "manage.py", "migrate"]
    x-init: true
  • Supports pre-start init containers, allowing services to declare initialization containers that run and complete before the main service starts.

1 detail in Everything

1 more release in this issue · 2026-06-23 → 2026-07-02
v5.2.0 Notes Stable

Docker Compose v5.2.0 introduces a new reconciliation algorithm and a rawsetenv message type for provider plugins.

└──▷ GET THIS VERSION
$ git clone --branch v5.2.0 https://github.com/docker/compose.git
# already have the repo? check out this version:
$ git checkout v5.2.0
  • New reconciliation algorithm aligns observed state with expected state more accurately during docker compose up and related operations.
  • Adds rawsetenv message type for provider plugins, enabling plugins to set environment variables without shell interpolation.

2 details in Everything

Was this useful?
◆  Runtime Security

Docker Sandboxes

Sources Release notes → v0.35.0 3 releases · 2026-06-17 → 2026-07-10 Notes LatestStable

Docker Sandboxes v0.35.0 adds SOCKS5SOCKS5A proxy protocol that relays TCP and UDP traffic through an intermediary server without inspecting it, letting cyber tools tunnel connections through firewalls or pivot hosts with minimal overhead. proxy support, policy inspection/check commands, secret import, and a daemon management CLI.

└──▷ GET THIS VERSION
$ git clone --branch v0.35.0 https://github.com/docker/sbx-releases.git
# already have the repo? check out this version:
$ git checkout v0.35.0
└──▷ TRY IT
Before running an agent task, verify that the current policy permits a specific outbound connection — catch misconfigurations without launching the full sandbox.
$ sbx policy check network --host api.example.com --port 443
Route all sandbox egress through a corporate SOCKS5 proxy (with credentials) so sandbox traffic flows through your org's inspection infrastructure.
$ DOCKER_SANDBOXES_PROXY=socks5://proxyuser:proxypass@proxy.corp.example.com:1080 DOCKER_SANDBOXES_NO_PROXY=localhost,127.0.0.1 sbx run
Migrate host env-var credentials (e.g. API keys) into the keychain so they are explicitly managed rather than auto-injected at runtime.
$ sbx secret import
  • Adds SOCKS5 upstream-proxy transport (socks5:// / socks5h://) for sandbox egress via DOCKER_SANDBOXES_PROXY, HTTP_PROXY, or HTTPS_PROXY, with optional auth and DOCKER_SANDBOXES_NO_PROXY exclusions.
  • New sbx policy inspect command to examine policy details.
  • New sbx policy check network command to test whether the current policy would allow a network access request before executing.
  • Revamped sbx policy ls with --wide, --source, and --decision filters for concise policy listing.
  • New sbx secret import imports credential env vars into the keychain; sbx secret ls now flags env-only and OAuth-shadowed entries. Host env vars no longer auto-inject at runtime — migration required.
+10 more show less
  • New sbx daemon command with start, stop, status, and log-level subcommands for daemon lifecycle management.
  • sbx inspect now surfaces the sandbox's kits, injected secrets, and sandbox information.
  • sbx rm now requires --force to delete an active session, preventing accidental teardown.
  • sbx kit add now recreates the sandbox container with the full kit set (preserving state) and applies the kit's network allow/deny rules and composed policy.
  • Kits can inject the user's Docker login token into requests to docker.com hosts via a credential with service sbx-login.
  • Re-attaching to a sandbox created with a custom --kit agent now works via sbx run --name <name> without re-passing --kit.
  • Enables virtiofs caching by default on all operating systems for faster filesystem performance (opt out with DOCKER_SANDBOXES_ENABLE_VIRTIOFS_CACHE=0).
  • Balanced network preset now allows VS Code domains, Azure Blob Storage (*.blob.core.windows.net), and dhi.io over HTTP.
  • Droid OAuth credentials are now proxy-managed, keeping real tokens on the host and out of the sandbox.
  • Faster sandbox startup via optimized TLS-proxy CA installation (merging into trust bundle instead of running update-ca-certificates), saving several hundred milliseconds.

15 details in Everything

└──▷ BREAKING ON UPGRADE
  • !Host env vars no longer auto-inject at runtime — use sbx secret import to migrate existing env-var-based secrets.
  • !sbx rm now won't delete an active session unless --force is passed.
  • !Installation is blocked on Windows versions older than Windows 11.
2 more releases in this issue · 2026-06-17 → 2026-07-10
v0.34.0 Notes Stable

Docker Sandboxes v0.34.0 adds a native SSH endpoint, kit source allowlisting, OCIOCIOpen Container Initiative standard, maintained by the Linux Foundation, defining the image format and runtime spec that container tools share so images built once run anywhere OCI-compliant. v2 streaming, and a new onboarding command.

└──▷ GET THIS VERSION
$ git clone --branch v0.34.0 https://github.com/docker/sbx-releases.git
# already have the repo? check out this version:
$ git checkout v0.34.0
└──▷ TRY IT
Enable the experimental SSH endpoint so agents or operators can shell directly into a running sandbox without a separate proxy.
$ sbx settings set feature.ssh true
Connect an interactive shell to a named sandbox over the native SSH endpoint once feature.ssh is enabled.
$ ssh <sandbox-name>@127.0.0.1 -p 2222
Allow kit installs from a private registry or Git source in addition to Docker Hub, to avoid the new allowlist breaking change.
$ sbx settings set kit.allowedSources ghcr.io/myorg/
  • New experimental native SSH endpoint in sandboxd: connect with ssh <sandbox-name>@127.0.0.1 -p 2222 (publickey auth, connect-to-create, interactive shell and exec).
  • New experimental sbx setup command imports agent credentials from environment variables for smoother first-time onboarding.
  • Adds OCI v2 kit artifact streaming that decompresses layers once to a cache directory and uses seek-based random access, reducing memory pressure during kit reads.
  • Restricts kit installs to an allowlist of sources (default: Docker Hub only); installing from other registries or Git URLs now fails unless explicitly permitted — breaking change.
  • Renames sbx policy set-default to sbx policy init; old name kept as a hidden deprecated alias.
+1 more show less
  • Restores published sandbox ports on restart, with automatic recovery from host-port conflicts by selecting a new host port.

6 details in Everything

└──▷ BREAKING ON UPGRADE
  • !Kit installs are restricted to an allowlist of sources defaulting to Docker Hub (docker.io/) only; installing a kit from another registry or a Git URL fails until you add its prefix with sbx settings set kit.allowedSources.
  • !The sbx policy set-default command is renamed to sbx policy init (the old name still works as a hidden deprecated alias, but is deprecated).
v0.33.0 Notes Stable

sbx v0.33.0 adds named sandbox identity, DNS-gated network policy, wildcard secrets, and Cursor OAuth support.

└──▷ GET THIS VERSION
$ git clone --branch v0.33.0 https://github.com/docker/sbx-releases.git
# already have the repo? check out this version:
$ git checkout v0.33.0
└──▷ TRY IT
Run two independent Claude Code sandboxes in the same workspace — useful when parallelising agent tasks without them colliding.
$ sbx run --name feature-a claude && sbx run --name feature-b claude
Re-attach to a named sandbox from a different directory without specifying an agent, useful when returning to a long-running session.
$ sbx run --name feature-a
Grant a custom secret to all subdomains of an internal service using wildcard host patterns.
$ sbx secret set-custom --host '*.internal.corp' --host '**.api.internal.corp' MY_API_TOKEN
  • New --name flag for sbx run identifies a sandbox independently of the working directory, enabling multiple named sandboxes in the same workspace and re-attachment from any directory.
  • sbx ls --json now reports a stable per-sandbox id field for scripting and automation.
  • Sandbox DNS resolution is now gated on network policy, closing a DNS-based data-exfiltration channel (CVE-2026-12039); loopback names remain exempt.
  • Outgoing ICMP from sandboxes is now blocked persistently across daemon restarts (CVE-2026-12539).
  • sbx secret set-custom --host and serviceDomains in kits now accept wildcard host patterns (* for one label, ** for any number) and are repeatable, allowing one secret to cover multiple subdomains.
+8 more show less
  • Adds Cursor OAuth support as a recognized agent.
  • virtiofs cache is now enabled by default on macOS and Linux, improving filesystem performance.
  • Linux arm64 build packages are now produced, extending platform support.
  • New keychain fallback to encrypted on-disk store on Linux when dbus-launch is unavailable, enabling headless/server deployments.
  • MITM proxy now publishes a CRLCRLA Certificate Revocation List is a signed file, published by a certificate authority, listing certificates it has revoked before expiry so that clients can reject compromised or mis-issued credentials without contacting the CA live. and embeds a CRL distribution point in generated certificates, supporting revocation-strict TLS clients (e.g., .NET).
  • Claude connectors (Slack, Gmail, Notion, Atlassian, etc.) now work inside sandboxed Claude Code without manual policy overrides.
  • sbx cp -L now follows symlinks in the source path for sandbox-to-host copies.
  • sbx exec now uses the same working directory as sbx run.

13 details in Everything

└──▷ BREAKING ON UPGRADE
  • !The sbx run <name> positional argument form for re-attaching to an existing sandbox now prints a deprecation warning; the positional argument for sbx run is expected to be an agent (e.g. claude, codex), not a sandbox name, and the old positional-name form will be removed in a future release.
  • !sbx run --name no longer auto-creates numbered sibling sandboxes when a same-named sandbox exists.
  • !sbx create now errors when run without arguments instead of proceeding.
Was this useful?
Detection & Response
◆  Endpoint Security (EDR/XDR)

Microsoft Sysinternals

Sources Blog / feed → LiveKd v5.65, ProcDump v12.01, and ZoomIt v12.11 (2026-07-09) Blog

Microsoft Sysinternals is a suite of advanced Windows system utilities for troubleshooting, monitoring, and diagnosing operating system internals, processes, and security issues.

LiveKd 5.65 restricts dump file access to admins; ZoomIt 11.11 adds JPEG and WEBP snip format support.

  • LiveKd now protects generated kernel dump files from being read by non-admin users.
  • ZoomIt adds JPEG and WEBP format support for snip captures.

2 details in Everything

Was this useful?
◆  Detection Engineering

Datadog Stratus Red Team

Sources Release notes →Source code → v2.34.0 Notes Code

Granular, Actionable Adversary Emulation for the Cloud

Stratus Red Team v2.34.0 adds Azure App Service credential harvesting, a controllable logger API, and a threat-intel coverage mapping skill.

└──▷ GET THIS VERSION
$ git clone --branch v2.34.0 https://github.com/DataDog/stratus-red-team.git
# already have the repo? check out this version:
$ git checkout v2.34.0
└──▷ USE IT
Route all Stratus library log output through your own zap logger instead of the default stdout handler — useful when embeddingembeddingLists of numbers a model produces to represent the meaning of text, images or code, so similar items sit close together. They power semantic search and RAG. Stratus in a larger tool.
go
import (
    "log/slog"
    stratuslog "github.com/datadog/stratus-red-team/v2/pkg/stratus/log"
    "go.uber.org/zap"
    "go.uber.org/zap/exp/zapslog"
)

func init() {
    zapLogger, _ := zap.NewProduction()
    handler := zapslog.NewHandler(zapLogger.Core(), zapslog.WithCaller(false))
    stratuslog.SetLogger(slog.New(handler))
}
  • New Azure attack technique: Retrieve App Service Publishing Credentials (azure.credential-access.app-service-publishing-credentials).
  • New pkg/stratus/log package exposes a controllable logger, letting library embedders route Stratus log output through any slog.Handler (e.g., zap, logrus).
  • New map-threat-intel-coverage agent skill fetches a threat-intelligence report, extracts cloud TTPsTTPsTactics, techniques and procedures, the goals, methods and specific behaviors used by an attacker. They describe how an attack works, rather than only the files or addresses it leaves behind., classifies each against existing Stratus Red Team coverage, and drafts GitHub issues for gaps.

3 details in Everything

Was this useful?
Governance, Risk & Compliance
◆  Compliance Automation & Trust Management

Vanta

Sources Release page → June 2026 Notes

Vanta June 2026 adds audit commenting, 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. expansion, AIUC-1 pilot, CrowdStrike CSPMCSPMCloud security posture management, checking cloud resources for risky settings and policy violations. It helps teams find exposed services, overly broad permissions and other configuration gaps., Twilio integration, and 20 new privacy MCP tools.

└──▷ HOW TO FIND IT
Discuss a specific control gap with your auditor in-context during an audit, without leaving Vanta.
📍In the Vanta console, go to Audits › [your audit] › Controls, click a control row, select Comments, type your message, and @mention your auditor to notify them directly.
Connect a custom or internal MCP-capable tool to Vanta's MCP to automate compliance workflows without waiting for manual allowlisting.
📍In the Vanta console, go to Integrations › MCP, copy the MCP connection URL, paste it into your MCP-capable tool. If the redirect URL is not Vanta-owned, acknowledge the warning banner to proceed.
Ingest CrowdStrike CSPM cloud misconfiguration findings into Vanta to track remediation SLAs and generate audit evidence.
📍In the Vanta console, go to Integrations › CrowdStrike, select Add Connection, choose your CrowdStrike tenant, and enable CSPM Alerts ingestion. Repeat for additional tenants under the same Vanta domain.
  • Adds threaded, @mention-enabled commenting on individual controls directly within a Vanta audit, replacing out-of-band email and Slack threads.
  • Enables one-click export of the full controls table from within a Vanta audit for direct use in audit reports.
  • Expands Vanta MCP connectivity to any MCP-capable tool (not just officially supported ones), with a warning banner when the redirect URL is outside Vanta-controlled origins.
  • Adds 20 new privacy MCP tools covering data inventory and privacy assessments, enabling agentic automation of privacy program tasks.
  • Introduces early-access pilot for AIUC-1, the new security standard for AI agents, available at no cost during the preview period.
+7 more show less
  • Adds Azure DevOps multi-tenant support, allowing multiple Azure DevOps organizations to be connected and managed under a single Vanta domain.
  • New Twilio integration (OAuth-based) automatically imports active Twilio users for access reviews and audit-ready evidence, with AI-summarized activity log.
  • Adds CrowdStrike CSPM ingestion into Vanta Security Alerts for SLASLAA formal agreement between a service provider and customer that defines expected uptime, response times, and remedies for failures, giving cyber tools a contractual baseline for availability and support commitments. tracking and audit evidence, plus multi-tenant CrowdStrike support.
  • Adds optional reason field when deactivating controls (individually or in bulk), surfaced in a new Deactivation Reason column and control history.
  • Upgrades AI evidence evaluation model to Gemini 3.5 Flash, delivering a 10% latency improvement and 5% higher evidence pass rate.
  • SOC 2SOC 2An auditing framework from the American Institute of CPAs that certifies a service organization's controls around security, availability, and data handling, giving buyers assurance that a vendor meets baseline trust criteria. Guide users can now have the Vanta Agent automatically generate a Trust Center and Letter of Engagement pre-filled with key details within their first days.
  • Unmanaged auto-renewal users can now preview a full line-item bill breakdown on the billing page before renewal processes.

12 details in Everything

Was this useful?
◆  Integrated & IT Risk Management

LogicGate

Sources Blog / feed → GRC Has Entered Its Agentic Era. Here’s What That Means For Your Team. (2026-06-30) 2 releases · 2026-06-11 → 2026-06-30 Blog LatestStable

LogicGate Spring 2026 introduces seven AI Workflow Agents, AI Governance module, expanded Spark AI, and enterprise-wide search across GRCGRCGovernance, Risk, and Compliance — a framework for aligning IT controls with business policy, regulatory requirements, and risk management so organizations can audit and report on their posture systematically. use cases.

  • Launches seven new AI Workflow Agents in early access spanning Third-Party Risk ManagementThird-Party Risk ManagementAssessing and monitoring the security risk an organization takes on through its vendors, suppliers and service providers. It covers the questionnaires, evidence and continuous checks behind letting a third party hold your data., Enterprise Risk Management, Business Continuity Management, and AI Governance use cases.
  • New AI Governance module provides a unified view of AI risk: maps the full AI ecosystem, auto-calculates risk levels using 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. categories, and surfaces business-impact indicators aligned to 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. and NIST frameworks.
  • Adds Reporting Insights to Spark AI, enabling analysis of large data volumes within a single report with guided next-action recommendations.
  • Introduces Automated Evidence Testing, which tests control evidence at the moment of collection to remove manual bottlenecks from the control lifecycle.
  • Expands Config Newton (agentic GRC engineer) to work alongside in-house teams, enabling application deployment 50–75% faster plus bespoke reporting and end-user training assistance.
+3 more show less
  • New App-Agnostic Record Search lets practitioners instantly find any Record they have access to across all Risk Cloud applications.
  • New Sticky Application Switcher enables rapid context-switching between risk domains without losing workflow state.
  • New Smart Automations autonomously transition workflow steps and send custom alerts based on configurable trigger conditions.

8 details in Everything

1 more release in this issue · 2026-06-11 → 2026-06-30
The Agentic Era, Live: A Recap of Agility 2026 (2026-06-11) Blog Stable

LogicGate unveils Config Newton agentic GRCGRCGovernance, Risk, and Compliance — a framework for aligning IT controls with business policy, regulatory requirements, and risk management so organizations can audit and report on their posture systematically. engineer and four new workflow agents for TPRMTPRMAssessing and monitoring the security risk an organization takes on through its vendors, suppliers and service providers. It covers the questionnaires, evidence and continuous checks behind letting a third party hold your data. and AI governance.

  • Introduces Config Newton, an agentic GRC engineer that interprets complex program requirements, navigates Risk Cloud, and autonomously builds complete GRC programs.
  • New Third-Party Risk ManagementThird-Party Risk ManagementAssessing and monitoring the security risk an organization takes on through its vendors, suppliers and service providers. It covers the questionnaires, evidence and continuous checks behind letting a third party hold your data. Agents automatically route vendors based on risk signals, complete formal assessments, and flag findings for human-led remediation.
  • New AI Governance Agents triage incoming AI use cases, score them against your risk framework, and surface recommended mitigations for human review.
  • All agent actions are fully logged and auditable, keeping practitioners in the decision loop while automating routine workflow progression.

4 details in Everything

Was this useful?
Build
◆  AI Coding Agents

Claude Code

Sources Release notes →Source code → v2.1.207 Notes Code

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.

Auto mode goes GA on Bedrock/Vertex/Foundry, Opus 4.8 becomes the default, and plugin config scoping tightens.

└──▷ GET THIS VERSION
$ git clone --branch v2.1.207 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.207
└──▷ USE IT
Opt out of auto mode on a Bedrock or Vertex deployment where you want to keep manual control over model selection.
json
# In ~/.claude/settings.json
{
  "disableAutoMode": true
}
Migrate a per-repo auto mode override to the correct user-level settings file so it is still respected after this release.
json
# Move autoMode from .claude/settings.local.json to ~/.claude/settings.json
{
  "autoMode": "auto"
}
Update a plugin hook that previously used shell-form ${user_config.*} to the new exec-form with $CLAUDE_PLUGIN_OPTION_<KEY> to avoid the shell-injection rejection.
json
# Before (now rejected):
# command: "my-hook --token ${user_config.api_token}"

# After — exec form with env var:
{
  "args": ["my-hook", "--token"],
  "env": { "TOKEN": "$CLAUDE_PLUGIN_OPTION_API_TOKEN" }
}
  • Auto mode is now generally available on Bedrock, Vertex AI, and Foundry without the CLAUDE_CODE_ENABLE_AUTO_MODE opt-in flag; disable it via disableAutoMode in settings.
  • Bedrock, Vertex, and Claude Platform on AWS now default to Claude Opus 4.8.
  • Auto mode config is now read exclusively from ~/.claude/settings.json; the repo-resident .claude/settings.local.json autoMode key is no longer honored (breaking change for teams using per-repo overrides).
  • Plugin option values (pluginConfigs) are no longer read from project-level .claude/settings.json; only user-level, --settings, and managed settings are honored (breaking change for project-scoped plugin configs).
  • Plugin hooks/monitors/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. headersHelper: ${user_config.*} interpolation in shell-form commands is now rejected; hooks must use exec form (args array) or $CLAUDE_PLUGIN_OPTION_<KEY> env vars instead (breaking change).
+2 more show less
  • /doctor now reports when the launcher at ~/.local/bin/claude is externally managed (e.g. a custom script or symlink).
  • Amounts over $1,000 in /usage-credits now require a typed confirmation, and malformed inputs are rejected with an error rather than silently stripped.

7 details in Everything

└──▷ BREAKING ON UPGRADE
  • !Auto mode no longer reads autoMode from .claude/settings.local.json (repo-resident); use ~/.claude/settings.json instead
  • !${user_config.*} in shell-form commands for plugin hooks/monitors/MCP headersHelper is now rejected; hooks must use exec form (args array) or $CLAUDE_PLUGIN_OPTION_<KEY>; monitors and headersHelper must read the value inside the script
  • !Plugin option values (pluginConfigs) are no longer read from project-level .claude/settings.json; only user, --settings, and managed settings are honored
Was this useful?

Pi

Sources Release notes → v0.80.6 7 releases · 2026-06-18 → 2026-07-09 Notes LatestStable

AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

Pi v0.80.6 adds a max thinking level for GPT-5.6/Claude and input-token pricing tiers for long-context cost accounting.

└──▷ GET THIS VERSION
$ git clone --branch v0.80.6 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.80.6
└──▷ TRY IT
Invoke the highest reasoning depth on a complex refactor task when using GPT-5.6 or a supported Claude model.
$ pi --thinking max 'Audit this codebase for insecure deserialization patterns and propose fixes'
  • Adds max thinking level (above xhigh) for GPT-5.6 and adaptive Claude models, available via CLI, SDK, RPCRPCRemote Procedure Call, a protocol that lets a program execute a function on another machine as if it were local, giving cyber tools a structured way to communicate across processes or hosts without managing raw sockets., and model selection; custom themes can define thinkingMax.
  • Adds request-wide input-token pricing tiers for accurate long-context cost accounting (e.g. GPT-5.4/5.5/5.6 rates), configurable in models.json and modelOverrides.
  • Supports ~ (home directory) expansion in the shellPath setting.

3 details in Everything

6 more releases in this issue · 2026-06-18 → 2026-07-09
v0.80.3 Notes Stable

Pi v0.80.3 adds Claude Sonnet 5, Azure Foundry endpoints, new RPCRPCRemote Procedure Call, a protocol that lets a program execute a function on another machine as if it were local, giving cyber tools a structured way to communicate across processes or hosts without managing raw sockets. session APIs, and configurable editor/padding settings.

└──▷ GET THIS VERSION
$ git clone --branch v0.80.3 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.80.3
└──▷ USE IT
Override the Ctrl+G external editor to use a specific binary (e.g. vim) instead of the environment default.
json
// In settings.json
{ "externalEditor": "vim" }
Tighten or expand terminal padding for assistant and thinking output to improve readability in narrow or wide terminals.
json
// In settings.json
{ "outputPad": 4 }
  • Adds Anthropic Claude Sonnet 5 support via Anthropic-compatible and Bedrock provider catalogs, with adaptive thinking enabled.
  • Supports modern Microsoft Azure Foundry endpoint URLs for the Azure OpenAI Responses provider.
  • New get_entries and get_tree RPC commands let clients inspect session entries and tree snapshots programmatically.
  • New ./rpc-entry package export enables launching Pi directly in RPC mode.
  • New session_info_changed extension event lets extensions observe session name changes.
+4 more show less
  • New externalEditor setting configures a preferred editor for Ctrl+G, overriding $VISUAL/$EDITOR fallbacks.
  • New outputPad setting controls horizontal padding for user messages, assistant messages, and thinking blocks.
  • Adds Usage.reasoning token counts for providers that report reasoning/thinking token usage.
  • Changes the default OpenAI model to gpt-5.5.

9 details in Everything

└──▷ BREAKING ON UPGRADE
  • !The default OpenAI model is changed to gpt-5.5.
v0.80.0 Notes Stable

Pi v0.80.0 adds Ctrl+J newline keybinding and migrates pi-ai's global API to a new compat entrypoint.

└──▷ GET THIS VERSION
$ git clone --branch v0.80.0 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.80.0
  • Adds Ctrl+J as a default newline keybinding alongside Shift+Enter.
  • Moves pi-ai's legacy global API (stream/complete/completeSimple, getModel/getModels/getProviders, registerApiProvider, getEnvApiKey, etc.) from the @earendil-works/pi-ai root to @earendil-works/pi-ai/compat; extension sources typechecking against pi-ai's published types should update imports accordingly or migrate to the new createModels()/provider-factory API.
  • Removes the @earendil-works/pi-ai/base and @earendil-works/pi-agent-core/base selective-provider entrypoints; use the root packages with explicit Models provider factories instead.

3 details in Everything

└──▷ BREAKING ON UPGRADE
  • !The @earendil-works/pi-ai/base and @earendil-works/pi-agent-core/base entrypoints have been removed; use the root packages with explicit Models provider factories instead.
  • !The global API (stream/complete/completeSimple, getModel/getModels/getProviders, registerApiProvider, getEnvApiKey, etc.) has been moved off the @earendil-works/pi-ai root entrypoint to @earendil-works/pi-ai/compat; extension sources that typecheck against pi-ai's published types must switch those imports to @earendil-works/pi-ai/compat or migrate to the new createModels()/provider-factory API.
v0.79.10 Notes Stable

Extension compaction events gain reason/willRetry context; pi update now pins the exact checked version.

└──▷ GET THIS VERSION
$ git clone --branch v0.79.10 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.79.10
└──▷ USE IT
React differently in a custom summarization extension depending on whether compaction was triggered manually, by a threshold, or as an overflow retry.
javascript
// In your extension handler:
extension.on('session_before_compact', ({ reason, willRetry }) => {
  if (reason === 'overflow' && willRetry) {
    // overflow retry: use a lighter summary to reduce token cost
  } else if (reason === 'manual') {
    // user-initiated /compact: run full archival summary
  }
});
  • Adds reason and willRetry metadata to session_before_compact and session_compact extension events, enabling extensions to distinguish manual /compact, threshold auto-compaction, and overflow retry flows.
  • Updates pi update to install the exact version returned by the Pi update check and surface the changelog URL in update notices, making upgrade behavior predictable.

2 details in Everything

v0.79.9 Notes Stable

Pi v0.79.9 adds chat-template thinking support for vLLM/Hugging Face providers like DeepSeek and improves GLM-5.2 routing.

└──▷ GET THIS VERSION
$ git clone --branch v0.79.9 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.79.9
  • Adds chat-template thinking support: OpenAI-compatible custom providers can now map Pi thinking levels into chat_template_kwargs, enabling vLLM/Hugging Face models such as DeepSeek to use provider-native thinking controls.
  • Enables GLM-5.2 xhigh reasoning on OpenRouterOpenRouterA unified API gateway that routes requests to dozens of LLM providers under one endpoint, letting cyber tools swap or compare models without rewriting provider-specific integration code. and corrects Fireworks routing to use the OpenAI-compatible Chat Completions endpoint with reasoning_effort support.

2 details in Everything

v0.79.8 Notes Stable

Pi v0.79.8 adds Mistral prompt caching, OpenRouterOpenRouterA unified API gateway that routes requests to dozens of LLM providers under one endpoint, letting cyber tools swap or compare models without rewriting provider-specific integration code. Fusion alias, and post-compaction token estimates.

└──▷ GET THIS VERSION
$ git clone --branch v0.79.8 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.79.8
  • Adds Mistral prompt caching using the Pi session ID as the cache key, with cached-token usage and cost accounting surfaced per session.
  • Adds openrouter/fusion as a built-in OpenRouter model alias for direct use as a provider target.
  • Adds estimated post-compaction token counts to compact results and compaction events, letting clients display approximate context reduction.
  • Adds @earendil-works/pi-ai/base and @earendil-works/pi-agent-core/base entry points for SDK users who need selective provider registration to trim bundle size.

4 details in Everything

v0.79.7 Notes Stable

Pi v0.79.7 adds automatic light/dark theme switching, scoped update control, and new extension API exports.

└──▷ GET THIS VERSION
$ git clone --branch v0.79.7 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.79.7
└──▷ TRY IT
Update pi and all extensions in one shot when you want everything on the latest version.
$ pi update --all
  • Adds automatic theme mode in /settings to assign separate light and dark themes that follow terminal color-scheme changes.
  • Changes pi update to update only pi by default; new pi update --all updates pi and all extensions together.
  • Exports generateDiffString, generateUnifiedPatch, and EditDiffResult from the public API for extensions that produce edit-style diffs.
  • Enables inline image rendering in Warp terminals via automatic Kitty graphics protocolKitty graphics protocolA terminal graphics protocol created by the Kitty terminal emulator project that lets applications render pixel-accurate images directly inside a terminal, giving cyber tools rich visual output without a GUI. detection.

4 details in Everything

└──▷ BREAKING ON UPGRADE
  • !The pi update command now updates only pi itself; the previous behaviour of updating pi and packages together now requires pi update --all.
Was this useful?

Cline

Sources Release notes →Source code → v4.0.8 Notes Code

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

GCP Vertex provider gains new Claude models and a free-form model ID entry field.

└──▷ GET THIS VERSION
$ git clone --branch v4.0.8 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v4.0.8
  • Adds new Claude model IDs to the GCP Vertex provider, including claude-sonnet-5, claude-opus-4-x variants, claude-haiku-4-5, and claude-fable-5.
  • Enables free-form model ID entry in the Vertex provider's model dropdown, letting practitioners use any Vertex-hosted model without waiting for an extension release.

2 details in Everything

Was this useful?

OpenHands

Sources Release notes → cloud-1.46.0 Notes

OpenHands: AI-Driven Development

OpenHands 1.46 adds configurable BYOR key alias patterns and enriched final archive manifests.

└──▷ GET THIS VERSION
$ git clone --branch cloud-1.46.0 https://github.com/OpenHands/OpenHands.git
# already have the repo? check out this version:
$ git checkout cloud-1.46.0
  • Supports configurable Bring-Your-Own-Router (BYOR) key alias patterns for enterprise deployments.
  • Enriches final archive manifests with additional metadata and removes redundant initial snapshots.

2 details in Everything

Was this useful?

Cursor

Sources Release page → Side Chats and Conversation Search Notes

Cursor is an AI-powered code editor that helps developers write, edit, and debug code faster using integrated large language model assistance.

Cursor adds side chats, agent transcript search, new cloud agent hooks, and redesigned project/repo pickers.

└──▷ HOW TO FIND IT
Spin up a side chat to sanity-check a design decision while the main agent keeps running — without losing your place.
📍In the chat panel, type /side or /btw to open a new side chat that already has context from your main conversation.
Find a specific past agent conversation by searching full transcripts, not just names or PR numbers.
📍In the Agents Window, press Cmd+K and type keywords from the transcript content to locate the conversation.
Build a self-correcting agent loop by intercepting responses and controlling subagent behavior via new cloud agent hooks.
📍In your cloud agent hook configuration, register handlers for beforeSubmitPrompt, afterAgentResponse, afterAgentThought, stop, and subagentStart to observe reasoning and inject corrections. See supported hooks in the Cursor docs.
  • New Side Chats: open parallel agent conversations via /side, /btw, or the plus button without interrupting the main agent thread — each is a durable, full agent conversation that can be revisited and @-mentioned.
  • Adds agent transcript search in the Agents Window via Cmd+K (command palette), with a local search index that scales to thousands of conversations.
  • Adds in-conversation search with Cmd+F, match counter, and jump-between-matches for long agent transcripts.
  • New cloud agent hooks (beforeSubmitPrompt, afterAgentResponse, afterAgentThought, stop, subagentStart, and more) enable observing and controlling prompts, responses, reasoning, subagents, compaction, and turn completion.
  • Redesigned project and repo pickers support creating a project and connecting GitHub, GitLab, or Azure DevOps without leaving the picker.
+4 more show less
  • New Run on picker surfaces where an agent can execute — Cloud, This Computer, or Remote Machines — in a single unified selector.
  • Adds a Select Multiple toggle inside Cloud and This Computer flyouts for multi-repo and multi-root selection, replacing the separate Set Up Workspace builder.
  • Combines remote machines, team pools, and remote workspaces into one searchable Remote Machines menu.
  • Branch picker now opens on the default branch and recently used branches instead of a flat list.

9 details in Everything

Was this useful?

Windsurf

Sources Release page → updated · Jul 11, 2026 Notes

Windsurf is an AI-powered code editor that helps developers write, edit, and navigate code faster using intelligent autocomplete and agentic AI assistance.

└──▷ USE IT
Exclude a sensitive command (e.g., a secret-scanning script) from the sandbox so it runs with full environment access, enforced for the whole team.
json
{
  "sandbox": {
    "excluded": {
      "deny": ["run-secret-scan.sh"],
      "allow": ["git", "make test"]
    }
  }
}
Check how many ACUs your current Devin session has consumed without leaving the chat.
$ /usage
Suppress Devin attribution lines in commit messages made by the local agent — useful when commits must conform to org policy.
json
{
  "attribution": false
}
  • Skill permissions: frontmatter definitions now apply to auto-approvals.
  • Enterprise login policies are now enforced in the CLI.
  • Adds server-level 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 permission options: approve all tools on a server for the current session, or permanently.
  • Introduces Devin for Terminal — a CLI agent running locally in Rust with multi-model support and seamless hand-off to Devin Cloud.
  • Integrates Devin Cloud agent directly inside Windsurf, available with every self-serve plan, allowing one-click delegation to a cloud VM.
Was this useful?
Agent
◆  AI Agent Frameworks

LangGraph

Sources Release notes → cli==0.4.31 Notes

Build resilient agents.

LangGraph CLI now supports prebuild images for langgraph deploy.

└──▷ GET THIS VERSION
$ git clone --branch cli==0.4.31 https://github.com/langchain-ai/langgraph.git
# already have the repo? check out this version:
$ git checkout cli==0.4.31
Was this useful?
browse all tools → score card for your toolchain →