The Claude Code Changelog
It feels like every day there's a new Claude Code update. This unofficial podcast, hosted by Lachlan and James (& powered by Jellypod AI), helps us stay up to date with everything Anthropic ships to Claude, from product updates, new workflows, and more. Each episode highlights what changed, how to use it, and why it matters.
All Episodes
Latest Episodes
Claude Code’s New Auto-Backgrounding and Workflow Upgrades
We break down Claude Code’s new automatic MCP backgrounding for long-running tool calls, including the default 120,000 ms threshold, how to customize it with CLAUDE_CODE_MCP_AUTO_BACKGROUND_MS, and how setting it to 0 keeps workflows synchronous. The episode also covers handy quality-of-life updates like vim escape remaps, hidden working directories in startup banners, and live elapsed-time counters for active tool calls.Show NotesWeek 29 · July 13–17, 2026 - Claude Code Docs: https://code.claude.com/docs/en/whats-new/2026-w29
Securing Claude Code with a Process Wrapper
This episode breaks down how Claude Code’s process wrapper can keep enterprise AI tooling inside corporate security and telemetry controls, from launcher scripts and audit logging to avoiding runaway background daemons.It also covers practical gotchas like recursion loops, file permissions, VS Code spawning behavior, and a few handy updates such as Vim remaps and clearer MCP server errors.Show NotesClaude Code changelog - Claude Code Docs: https://code.claude.com/docs/en/changelog
Claude Code’s Windows Backslash Bug
This episode breaks down a Windows bug in Claude Code where unescaped backslashes in file paths triggered Unicode parsing errors, causing crashes and garbled folder names. It also covers recent Windows-focused fixes, including safer Git Bash path handling and a PowerShell permission patch, with a reminder to update to the latest release.
Claude Code 2.1.219: Nested Subagent Streaming Unlocked
We break down how Claude Code v2.1.219 finally exposes nested subagent activity in headless mode, making it easier to debug streaming text, thinking blocks, and tool events in real time. The episode also covers new JSON fields, MCP init error reporting, and practical guardrails for keeping parser pipelines fast and costs under control.
Claude Code’s DirectoryAdded Hook for Monorepos
This episode explores how Claude Code’s DirectoryAdded hook helps monorepos dynamically load the right environment for newly registered subdirectories, avoiding toolchain mix-ups like using npm in a bun workspace. It also covers workspace trust, hook exit codes, and a few quality-of-life fixes in the latest release.
Lock Down Claude Code with Strict Network Allowlists
Learn how Claude Code’s strict allowlist mode can help prevent prompt-fatigue mistakes by blocking unapproved network access by default. The episode also covers the config setup, common breakages like package installs, and why enterprise teams may still need a corporate proxy for deeper TLS inspection.
Claude Opus 5 Brings a Million-Token Context to the CLI
We break down the latest Claude Code update, where Opus 5 becomes the default model and a native one million token context window changes how developers can work with huge codebases. The episode also covers the new pricing model, prompt caching benefits, and quality-of-life improvements like stricter sandbox network controls and better Vim mode behavior.
Claude Code v2.1.219: Depth 3 and Safety Valves
The hosts unpack Anthropic’s return to depth 3 subagent nesting in v2.1.219, along with the new workflowSizeGuideline soft limit for keeping agent sprawl in check. They also cover practical CLI improvements like the network strict allowlist, the DirectoryAdded hook, and bug fixes for GNU screen and Vim mode.
Claude Code v2.1.218: Background Skills and Safer Workflows
This episode breaks down Claude Code v2.1.218’s new context: fork behavior, which runs custom skills asynchronously so you can keep coding while background analysis finishes. It also covers workspace trust protections, MCP diagnostics, cleaner agent naming, and a new prompt safeguard against accidental history wipes.
Claude Code v2.1.218: Background Reviews and Key Fixes
This episode breaks down the latest Claude Code release, including how code reviews now run in an isolated background subagent to avoid bloating your prompt history. It also covers the breaking change for forked skills, Windows path and left-arrow fixes, plus richer MCP debugging and config linting improvements.
Claude Code v2.1.217 Stops Runaway Subagents
We dig into Anthropic’s Claude Code v2.1.217 update, which slams the brakes on runaway nested subagents and adds real-time budget enforcement to stop surprise costs mid-task. Plus, a quick look at quality-of-life tweaks like emoji autocomplete, clickable SSH links, and disk space warnings.
Claude Code v2.1.216: Safer Dev Containers, Faster Resumes
This episode digs into Claude Code v2.1.216, including the new sandbox.filesystem.disabled setting that decouples filesystem isolation from network protections for safer dev-container workflows. It also covers faster resume times after quadratic slowdown fixes and a more reliable AskUserQuestion flow that reduces unsafe agent drift.
Cloud Swarms and Async AI Code Reviews
Explore how cloud-based ultrareview offloads AI code reviews to a multi-agent swarm, handling branches and PRs asynchronously without clogging your local terminal. The episode also covers recent fixes for merge-base edge cases, smarter PR parsing, and tighter billing safeguards for premium reviews.Show NotesUltrareview for Claude Code: Multi-Agent Cloud Code Review ...: https://claudedirectory.org/blog/ultrareview-claude-code-guide
Claude Code Reins In Rogue Agent Drift
We break down the latest Claude Code update that reins in autonomous behavior, making /verify and /code-review fully manual again. The episode also covers new workflow controls like /code-review --fix, remote reviewer fleets, progress heartbeats, and a Linux bug fix that stopped the CLI from self-destructing.Show NotesClaude Code changelog - Claude Code Docs: https://code.claude.com/docs/en/changelog
Claude Code v2.1.214 Locks Down Bash and Docker
This episode breaks down the latest Claude Code safety patch, including tighter Bash command analysis, stricter directory scoping, and new fail-closed prompts for risky redirections and obfuscated commands.It also covers Docker daemon lockdowns, the new EndConversation safeguard for prompt injections, and a Linux fix that prevents pkill -f from killing Claude itself.
Claude Code 2.1.212: Forking Terminal Workflows
We unpack Claude Code v2.1.212’s new parallel workflow tools, including /fork for detached background sessions and /subtask for inline delegation. The episode also covers new safety limits, automatic backgrounding for slow MCP tool calls, and the revamped /resume session picker for managing multiple agent runs.Show NotesClaude Code changelog - Claude Code Docs: https://code.claude.com/docs/en/changelog
Terminal Spoofing, Safety Guardrails, and Claude Code v2.1.211
We dig into a sneaky terminal spoofing exploit involving hidden Unicode characters and how Claude Code v2.1.211 adds visual sanitization to keep what you see aligned with what actually runs. The episode also covers hardened auto-mode boundaries, subagent text forwarding for auditing, and fixes for session cost tracking and Vim keybindings.
Claude Code v2.1.210: Unified Permissions and Better Terminal UX
This episode breaks down Claude Code v2.1.210’s simplified permission model, including the move from fragmented rules like Write and NotebookEdit to unified Edit and Read settings. It also covers handy terminal improvements like live tool timers, more reliable session attachment, and new Vim insert-mode remaps.
Claude Code v2.1.209: Vim Remaps, Screen Reader Mode, Bug Fixes
We break down the latest Claude Code updates, including vim-insert-mode remaps for faster terminal editing, a cleaner screen reader mode, and fixes for token parsing and background agent permission prompts.
Claude Code Plugin Security: Fixing Shell Injection
This episode breaks down the breaking changes in Claude Code 2.1.207, including why raw template strings in custom hooks were removed and how shell-injection attacks can happen through plugin configs.It also covers the safer migration paths with exec-form arrays and CLAUDE_PLUGIN_OPTION_ environment variables, plus the new decision to ignore local repository config files for plugin resolution.
Claude Code Locks Down Auto Mode and Respects Your Launcher
This episode covers a major security change in Claude Code 2.1.207 that moves Auto Mode permissions out of repo-local configs and into the global user settings file to prevent malicious workspace hijacking. It also highlights launcher protection improvements, smarter doctor checks, better autocomplete for directory navigation, and a fix for push behavior that now respects your Git remote settings.
Claude Code 2.1.207: Auto Mode and Security Fixes
This episode breaks down Claude Code 2.1.207, including Auto Mode now enabled by default on major enterprise platforms and the shift to safer, global-only config handling. It also covers plugin injection protections, better path autocomplete, and fixes for AWS SSO, Windows credential stalls, and terminal lag.
Trim CLAUDE.md and Speed Up Claude Code
Claude Code v2.1.206 adds a smarter /doctor check to flag bloated CLAUDE.md files that slow down prompts, waste tokens, and break caching. The episode also covers improved /cd autocompletion, safer git worktree writes, push-default handling, and a fix for resume/continue input bugs.
Claude Code 2.1.205: Self-Healing Checkups and Auto Mode Security
This episode covers Claude Code v2.1.205’s self-repairing /doctor and /checkup commands, which fix shell integrations and clean up stale local tools automatically. It also dives into major Auto Mode hardening, including fabricated approval blocking, safer deletions, better session warnings, lower-memory updates, and a fix for the dangerous NTFS junction delete bug.
Claude Code v2.1.204 Fixes MCP Sync, Hooks, and Memory Stalls
We break down the newest Claude Code fixes for dynamic MCP root syncing, including real-time notifications/roots/list_changed updates that avoid session restarts in monorepos and worktrees.Also covered: streaming stdout for headless SessionStart hooks, macOS stall fixes, Windows auth environment inheritance, and a leaner context-usage indicator that cuts CPU burn.
Claude Code v2.1.202 Splits Fast Reviews from Deep Agent Checks
This episode breaks down Claude Code’s new review workflow split, where /review is now a quick pass and /code-review handles the full multi-agent analysis, fixes, and PR comments. It also covers confidence thresholds, severity labels, cost considerations, and smaller quality-of-life updates like better SSH sign-in links and OpenTelemetry tags.
Claude Code v2.1.200: Manual Mode and Timeout Pitfalls
This episode breaks down the safety changes in Claude Code v2.1.200, including the shift to Manual permission mode and the new behavior of AskUserQuestion in automated workflows. It also covers idle timeout configuration for CI jobs, tmux 3.4 rendering fixes, and improved installer diagnostics for out-of-memory failures.
Claude Code Gets Dataviz, Stacked Skills, and Better Resilience
This episode explores Claude Code v2.1.198 and v2.1.199, including the new /dataviz skill for programmatic WCAG contrast checks and the ability to stack up to five slash-commands in one pipeline. It also covers improved network resilience, with smarter TLS retry handling, clearer proxy diagnostics, and preserved partial code streams during connection drops.
Claude Code Gets Safer Defaults and Better Cache Shields
This episode covers Claude Code v2.1.200’s shift to Manual permission mode, reducing the risk of unattended actions, and v2.1.201’s prompt-caching fix for Sonnet 5 that improves latency and cuts API costs.They also dig into the tmux flicker fix, synchronized terminal rendering, and a new installer warning that helps catch low-memory environments before a crash.
Claude Code v2.1.199 Fixes Stream Loss and TLS Hiccups
This episode covers Claude Code v2.1.199’s resilience upgrades, including preserved partial outputs after interrupted streams, higher retry limits for long-running agents, and a stabilized Linux daemon loop.It also digs into smarter SSL proxy diagnostics, the NODE_EXTRA_CA_CERTS fix for corporate TLS setups, and smoother chaining of slash commands.
Claude Code v2.1.198: Background Agents Go Fully Hands-Off
Claude Code’s latest release brings isolated git worktrees, automated commits, pushed changes, and draft PR creation to background agents so they can run without constant supervision. The episode also covers notification hooks, environment variable gotchas, automatic AWS token refresh, and improved retry handling for flaky network errors.
Claude Code v2.1.197: Sonnet 5 and the Million-Token Trap
This episode breaks down Claude Code v2.1.197, where Sonnet 5 becomes the default and a native 1M-token context window opens up new possibilities for giant repos and agentic workflows.We also cover the practical gotchas: prompt caching costs, cache-busting root config changes, recursive bash loops, and the latest quality-of-life fixes like clickable file paths, disabled mouse hijacking, and the new streaming watchdog.
Claude Code v2.1.196: Fixing RCE, Restoring 1M Tokens
We break down the security fix that pauses untrusted MCP servers with a new pending approval state, preventing auto-approved local code execution in cloned repos. Plus, a hidden bug that halves the auto-compact window, how to restore your full 1M-token context, and other quality-of-life updates and Linux regressions.
Claude Code Adds Agentic Bash Command Loops
This episode breaks down Claude Code’s new bash-command behavior, where failed terminal commands can automatically trigger a debugging loop, plus how to disable it when you want more control.We also cover workflow filtering, the new Skills section in /plugin, and a handy setting to stop terminal mouse clicks from hijacking your cursor.
Claude Code Fixes Background Daemons and RAM Hogs
This episode covers Claude Code’s new safeguards for unstable background daemons, including self-healing control sockets and automatic memory-pressure reaping in recent releases. It also explains how to opt out for long-lived local services and spotlights a small bash-mode shortcut for faster file path completion.
Claude Code v2.1.195: Exact-Match Hooks and Mouse Click Fixes
This episode breaks down Claude Code v2.1.195, including the shift from loose substring matching to exact-match security hooks for hyphenated MCP tools and the new wildcard syntax needed for multi-tool servers. It also covers the CLAUDE_CODE_DISABLE_MOUSE_CLICKS setting, plus smaller fixes for voice dictation and remote connections.
Claude Code v2.1.193: Telemetry, Security, and Shell Fixes
We break down the latest Claude Code update, including the new OpenTelemetry assistant response logging, the environment variable gotcha that can mirror user prompt settings, and the security boost from forcing every shell command through classification.Plus, a few quality-of-life improvements like file path autocomplete in interactive bash, and the new background shell reaping controls for keeping local dev machines stable.
Claude Code v2.1.191: Rewind, Rollbacks, and Faster Terminals
This episode explores Claude Code’s new /rewind feature for restoring code and conversation context after a /clear, plus the important limits of what it can and can’t undo.It also covers the release’s performance and usability upgrades, including a major CPU reduction in terminal streaming, smoother output handling, scroll-jump fixes, and more reliable background agent stops.
Claude Code Hardens Local Sandboxes and Remote Calls
This episode digs into Claude Code v2.1.187’s new sandbox.credentials setting, which blocks sandboxed access to local secrets like SSH keys, AWS files, and .env data while pushing workflows toward explicit, least-privilege access. It also covers the new five-minute timeout for remote tool calls and the MCP CLI login/logout commands that make secure remote integrations much easier to manage.
Headless Claude Code Auth and Safe Bash Automation
We break down how Claude Code handles MCP authentication on headless machines with no-browser login, making it much easier to use on remote servers and containers. The episode also explores its automatic shell execution safeguards and the clever iTerm2 Teammate Mode that keeps terminal workflows organized.
Stop Claude Code Push Spam with Presence Files
Learn how Claude Code’s CLAUDE_CLIENT_PRESENCE_FILE can quiet mobile notifications when you’re already at your desk, using lightweight screen-lock automation on macOS or Linux. The episode also covers faster startup with Bun 1.4, smoother line-by-line streaming, and handy inline config overrides.
Claude Code Fixes Premature Retries and Auto Mode Risks
This episode covers the latest Claude Code updates, including a longer stream-stall window that prevents premature retries and a friendlier waiting status for long-thinking tasks. It also highlights new Auto Mode safeguards that block destructive commands, plus a handy slash-config help command for easier setup.
Claude Code Teams, Nested Skills, and Privacy Fixes
The hosts explore Anthropic’s latest Claude Code CLI updates, including experimental agent teams, automatic nested skill discovery in monorepos, and cleaner sub-agent orchestration. They also cover the new /config workflow and a crucial privacy setting to remove tracking URLs from automated git commits.
Claude Code 2.1.183 Adds Guardrails for Rogue Commands
In this episode, Lachlan and James break down Claude Code 2.1.183 and its new protections against destructive Git, AWS, and Terraform commands in Auto Mode. They also cover the new attribution.sessionUrl option for cleaner commits and PRs, plus improved keyboard navigation in the config menu.
Claude Code 2.1.181 Adds Live /config Tweaks
We break down Claude Code’s latest release, including the new on-the-fly /config key=value command, faster mid-session tuning, and remote control support. The episode also covers mobile notification syncing with CLAUDE_CLIENT_PRESENCE_FILE and the new macOS sandbox option for Apple Events automation.
Claude Code v2.1.179 Fixes Drops, Globs, and WSL2
This episode covers Claude Code v2.1.179, including better handling of dropped socket connections, smarter sandbox glob processing, and reduced prompt bloat for large repositories.We also dig into quality-of-life fixes for WSL2 scrolling and faster remote plugin startup over SSH.
Claude Code v2.1.178: Parameter-Matched Permissions
We break down Claude Code v2.1.178’s new fine-grained permission system, including tool parameter matching like Tool(param:value) for blocking costly or risky actions before they run.We also cover smarter nested skill resolution, the new purple workflow indicator, improved /doctor diagnostics, and the stricter /bug submission flow.
Lock Down Claude Code Access with v2.1.176
Claude Code v2.1.176 tightens file access controls with corrected path specifiers, making rules like deny, ask, and allow work properly for sensitive files such as .ssh and .env. The hosts also cover handy tmux and SSH quality-of-life fixes, plus why this update is a must for safer terminal workflows.
Claude Code’s Hidden Footer Links and QoL Upgrades
We dig into undocumented gems in Claude Code v2.1.176, including footerLinksRegexes for clickable terminal badges, plus new quality-of-life tweaks like smoother scrolling controls and localized session titles.We also cover the improved AWS Bedrock credential caching that finally respects real token expiration for longer, less annoying dev sessions.
Claude Code v2.1.176 Fixes Hook Matching and Bedrock Caching
This episode breaks down Claude Code v2.1.176, including more reliable hook path matching for safer tool access, plus session title language matching that follows your conversation language.It also covers improved AWS Bedrock credential caching, making enterprise workflows smoother by reducing premature session interruptions.
Claude Code 2.1.174: Token Tracking and Model Guardrails
We break down the revamped /usage command in Claude Code v2.1.174, including detailed token attribution across skills, plugins, subagents, and cache misses. Then we cover v2.1.175’s enforceAvailableModels setting for tighter enterprise model control, plus a few handy terminal and exit-time fixes.
Claude Code’s 5-Level Agent Trees and Fable 5
We unpack Anthropic’s latest Claude Code update, where sub-agents can recursively spawn deeper agents, backed by a massive one-million-token context window. The conversation also covers token compaction, the new /plugin search bar, AWS Bedrock region fallback, and the risks of letting terminal automation grow into a full recursive agent tree.
Headless Claude Code: Agents, JSON Alerts, and Sandbox Locks
We dig into Claude Code v2.1.169’s new background-agent controls, including --json, --all, and the waitingFor field for spotting blocked automations. The discussion also covers enterprise hardening with CLAUDE_CODE_DISABLE_BUNDLED_SKILLS and the trade-offs between tighter security and losing built-in capabilities.
Claude Code v2.1.169: /cd, Safe Mode, and Faster Debugging
This episode covers the latest Claude Code update, including the new /cd command for switching directories without losing prompt cache and the safe mode option for isolating bugs caused by local plugins, rules, and hooks.It also touches on macOS rendering improvements, a Windows claude -p hang fix, and enterprise support for OpenTelemetry client certificates.
Claude Code Gets Version Gating for Enterprise Teams
We break down Anthropic’s new managed configuration scopes for Claude Code, including system-level version limits that keep teams on approved builds. Plus, we cover update UX improvements and safer slash-command autocomplete in the terminal.
Claude Code Hooks Let Agents Fix Their Own Failures
We break down Claude Code 2.1.163’s new hook behavior, where failed Stop and SubagentStop validations can feed compiler or linter output back into Claude for a self-correcting loop. The episode also covers new plugin list filters and a handy /btw clipboard shortcut for faster terminal workflows.
Claude Code Gets Fallback Models and Safer Defaults
We break down Anthropic’s latest Claude Code updates, including ordered fallback models for handling rate limits, new controls for disabling thinking tokens, and stronger permission prompts around sensitive system files.Also covered: the handy /btw clipboard command and why these small workflow improvements make agentic terminal work smoother and safer.
Claude Code v2.1.163: Hooks, TMPDIR, and Session Fixes
This update covers smarter lifecycle hooks that can return structured context directly to the model, making stop-phase workflows cleaner for testing and automation. It also fixes temporary directory handling for local builds and improves session ID propagation for stdio MCP servers when using resume and continue flows.
v2.1.162 Tightens WebFetch Security and Slash Command UX
This episode covers the latest v2.1.162 update, where WebFetch now honors explicit allowlists and blocklists over built-in domain rules for stronger control. It also highlights safer slash command autocomplete, read-only config fallbacks for containers, and the new waitingFor JSON key for better agent automation.
Claude Code v2.1.161: Fault-Tolerant Parallel Runs and OTEL Telemetry
This episode breaks down Claude Code v2.1.161’s new fault-tolerant parallel execution, where one failed tool call no longer kills the whole batch. The hosts also cover OpenTelemetry resource tagging, secret redaction in MCP listings, and improved terminal rendering in VS Code with the new /terminal-setup command.
Claude Code v2.1.160 Locks Down Shell Files
We break down the new permission safeguards in Claude Code v2.1.160, including explicit consent for writes to critical shell and config files like .npmrc and .zshenv. The episode also covers the faster read-before-edit check, token-saving grep behavior, and the rename of the dynamic workflow trigger to ultracode.
Claude Code v2.1.154: Dynamic Workflows and 16-Agent Parallelism
We break down Claude Code’s new Dynamic Workflows system, where a local JavaScript orchestrator replaces context-heavy multi-agent prompting for faster, cleaner execution. The episode also covers 16-agent parallel runs, the 1,000-invocation safety cap, Opus 4.8’s high-effort default, and the deprecation of the old fast-mode override in favor of /model and /fast commands.
Claude Code Auto Mode Hits Enterprise Clouds
We break down Claude Code 2.1.158’s new Auto Mode support for Bedrock, Vertex AI, and Palantir Foundry, and what it means for developers working inside tightly controlled enterprise environments.Plus: the new OpenTelemetry audit flag, full autonomous write-run-correct loops, and the fix for stubborn Git worktree locks.
Claude Code’s Local Skills: Speed vs Security
We break down Claude Code v2.1.157’s local .claude/skills workflow, where executable scripts can be loaded straight from a developer’s machine for instant tool creation and hot reloading. The conversation also digs into the security risks of committing custom plugins to source control, including the possibility of arbitrary code execution.
Claude Code Adds Auto-Fix and Simplify Cleanup
We dig into Claude Code’s new /code-review --fix workflow, which can edit files directly instead of just leaving comments, and the revamped /simplify pass for trimming dead code and redundant structure. The episode also covers safer usage tips, from keeping a clean git tree to reviewing diffs, plus the new skipLfs option for speeding up large repos.
Claude Code 2.1.153 Fixes Gateway Leak and Tightens MCP
We break down a critical credential leak in custom API gateways, where Claude Code mistakenly sent primary Anthropic OAuth credentials to the wrong place. The episode also covers faster git and GitHub plugin loading with skipLfs, plus new safeguards for subagent MCP configs to prevent privilege escalation.
Claude Code Fixes a 10GB MCP Leak and a Sneaky Resume Bug
This episode digs into the v2.1.132 Claude Code update, including a stdio MCP memory leak that could balloon long sessions past 10GB, plus the practical differences between stdio and HTTP transport. It also covers new session-aware Bash env vars, a terminal escape hatch for alternate screen mode, and a permission-mode bug that could quietly override your intent when resuming plan-mode sessions.
Claude Code 2.1.129 Turns Plugins Into Shareable Artifacts
This episode breaks down how Claude Code v2.1.129 changes plugin loading with URL-based zip installs, moving plugins from local setup to a true distribution model. It also explores the new experimental themes and monitors fields, and the security tradeoffs that come with treating plugins like packaged software.
Claude Code Fixes a Silent Worktree Bug
This episode breaks down a subtle Git behavior in Claude Code v2.1.128 where new worktrees now branch from local HEAD instead of origin/default-branch, preventing unpushed commits from being skipped. It also covers related reliability improvements, including OTEL isolation for subprocesses and better handling of parallel tool calls.
Claude Ultrareview Turns AI Feedback Into a Merge Gate
This episode explores how Claude Code’s `ultrareview` command uses exit codes, streamed output, and PR targeting to make AI review behave like a real CI check. It also digs into the multi-agent verification model, cost tradeoffs, and the network constraints teams need to consider before wiring it into GitHub Actions.
Anthropic’s xhigh: Claude Code’s New Default Gear
We break down Anthropic’s new xhigh effort level for Claude Code on Opus 4.7, why it now defaults for Max subscribers, and how the new slider and config flow make it easier to use.The discussion also covers when deeper reasoning is worth the extra latency and cost, plus how to spot the tasks where xhigh really pays off.
Claude Code v2.1.126 Finally Trusts Your Model Gateway
Claude Code now asks your configured ANTHROPIC_BASE_URL for /v1/models, so the model picker reflects the IDs your gateway actually exposes instead of a hardcoded Anthropic-only list. The episode breaks down why that matters for LiteLLM, Bedrock, and other proxies, plus the catch: the backend still has to return truthful model names or the client will pass them through verbatim.
Claude Code Gets a Clean Purge and Safer OAuth Login
We break down the new claude project purge command, which wipes stale transcripts, tasks, file history, and project config so broken project state can be reset cleanly. The episode also covers the expanded behavior of --dangerously-skip-permissions, the new direct OAuth code paste flow for login failures, and a subtle policy bug fix affecting Claude Code memory.
Claude Code’s Hidden Bedrock Cost Lever
April 29, 2026 brought a small Claude Code update with big budget implications: the new ANTHROPIC_BEDROCK_SERVICE_TIER setting lets Bedrock users choose default, flex, or priority and sends that choice as the X-Amzn-Bedrock-Service-Tier header. The episode breaks down when each tier makes sense, why latency and throughput guarantees matter, and how to avoid overpaying for interactive and batch workloads.
Claude Code v2.1.121: Instant MCP Tools and Tool-Output Hooks
This episode digs into two major v2.1.121 changes in Claude Code: the alwaysLoad flag for loading selected MCP server tools at session start, and a widened PostToolUse hook that can rewrite outputs from built-in tools like bash and file reads. It also covers the practical tradeoffs around context, trust boundaries, and the release’s memory leak fixes.
Claude Code Ultrareview Goes CI-Native
We break down how ultrareview moved from an interactive slash command to a real CLI that can run in scripts, Make targets, and GitHub Actions as a potential build gate. The episode also covers JSON output for automation, the trust and latency tradeoffs of model-based PR checks, plus smaller quality-of-life updates like PowerShell fallback on Windows and better GitHub attribution.
Claude Code Goes Beyond GitHub
This episode breaks down Claude Code v2.1.119’s expanded support for GitLab, Bitbucket, and GitHub Enterprise PRs, plus the value of loading full diff, commit, and description context automatically. It also covers smarter PR link handling, configurable review URL templates, and a small privacy win for demos with hidden working directories.
Claude Code’s 1-Hour Prompt Cache Upgrade
Claude Code v2.1.108 extends prompt caching from 5 minutes to 1 hour, making long system prompts and big repo contexts stay warm through real work sessions instead of expiring during normal breaks. The episode also covers the `DISABLE_TELEMETRY` bug fix, the optional 5-minute fallback, deprecated Bedrock compatibility, and why any prefix change still invalidates the cache immediately.
When /config Became Real Config in Claude Code
The hosts unpack how Claude Code 2.1.119 moved /config changes into ~/.claude/settings.json, making theme, editor mode, and verbose output persist across restarts. They also break down the config hierarchy across user, project, and managed policy scopes, plus what teams should audit after upgrading.
Claude Code Hooks Now Call MCP Tools Directly
This episode explores how Claude Code 2.1.118 and 2.1.119 turn hooks into first-class automation, letting PostToolUse events call MCP tools like Slack without wrapper scripts or brittle bash glue. It also digs into duration_ms as a clean timing signal for smarter logging, alerts, and workflow routing.
Claude’s Weird Week and Opus 4.7’s API Trapdoors
This episode digs into Anthropic’s postmortem on Claude’s silent output corruption, from misrouted traffic to TPU/compiler bugs that caused garbled or truncated responses for a significant share of users. It also covers Opus 4.7’s breaking API changes, including hard failures on familiar generation settings and a new tokenizer that can quietly raise token costs.
Claude Code Fixes a 5x Context Window Bug
This episode breaks down a compaction bug that made Claude Code treat Opus 4.7 like it had a 200K-token limit instead of 1M, causing long coding sessions to summarize far too early. It also covers the new default high effort setting for Pro and Max users on Opus 4.6 and Sonnet 4.6, and what that means for reasoning quality and token usage.
Claude Code 2.1.116: Faster Resume, Safer rm Rules
This episode breaks down Claude Code 2.1.116’s biggest changes: a much faster /resume flow for huge sessions, plus a critical safety fix that stops broad auto-allow rules from bypassing dangerous-path checks.We also cover hook behavior in --agent mode, MCP startup improvements, and a new GitHub rate-limit hint that makes automation feel less flaky.
Claude Code Adds xhigh: The New Middle Gear for Devs
The hosts dig into Claude Code 2.1.111’s new xhigh effort tier for Opus 4.7, explaining how it fills the gap between high and max for tougher coding and reasoning tasks. They also cover the redesigned /effort control, safe fallback behavior in mixed-model workflows, and the removal of the old auto-mode flag.
Cloud Code Review, Reproduced and Verified
The hosts unpack a cloud-based multi-agent code review command that runs in the background, checks either your current branch or a GitHub PR, and returns only independently reproduced findings. They also dig into why the feature feels more trustworthy than a single AI pass, plus the practical details of diffstat previews, faster launch checks, and how to avoid wasting review runs.
Claude Code 2.1.113: Faster Launches, Safer Bash Rules
This episode breaks down a small-looking Claude Code update that swaps in a native binary for faster cold starts and a leaner CLI, especially in CI. It also digs into important Bash rule fixes, including a simple env bypass, dangerous find -exec/find -delete approvals, and tighter handling of macOS private paths.
Claude Code Routines: Cloud Jobs, GitHub Triggers, Human Review
We break down Anthropic’s Claude Code Routines research preview, where jobs run in the cloud so your laptop can sleep, and explore the three ways to trigger them: schedules, API calls, and GitHub events.We also dig into the safety model behind the feature, from claude/ branches and repo scoping to the shift from autonomous execution to human review.
Claude Code 2.1.110: No More Flicker, Better Focus, Safer Edits
This episode digs into the latest Claude Code 2.1.110 improvements, including the new fullscreen TUI that stops terminal flicker without restarting your session and the cleaner split between /focus and Ctrl+O.It also covers practical upgrades like disabled auto-scroll, editor context injection, and important hardening changes that make long-running workflows and file handling more reliable.
Claude Code 2.1.108: Smarter Prompt Caching and Session Recaps
This episode breaks down Claude Code 2.1.108’s new prompt-caching options, including the tradeoff between longer-lived session continuity and short-lived cache turnover for testing. It also covers the new /recap workflow and background session summaries that make it easier to jump back into long-running coding work.
Claude Code Changelog, Decoded
Get a plain-English take on the latest Claude Code updates, with a focus on what changed, who it helps, and whether it actually matters. We separate the useful workflow tweaks from the noise so you can decide what to try, test, or ignore.