Telexed

AI news through a solo-operator lens — only what changes your day9 of 412

FILTER[All][Agents & tools][Models & API][Generative media][Infra & SaaS][ASO & growth][Indie business][Idea signals][Other][★6+ high-signal]

Claude Code Releases ✕clear filters

Yesterday1 dispatches

#0412
#0412Agents & tools Claude Code Releases17 hours ago
`Claude Code` `v2.1.144` improves background sessions, MCP tools, and terminal stability
80radar
Claude CodeCoding agent CLI — automates code work with Claude in terminal
Background agents are easier to resume and diagnose. The bigger win is fewer broken long sessions: MCP pagination, terminal corruption, startup hangs, and bad image files now fail less often.
- /resume now lists sessions started with claude --bg or agent view, marked as bg; background work is easier to recover after context switching.
- Background subagent completion notifications include elapsed time like 3h 2m 5s, useful for spotting expensive or slow automation runs.
- /model now changes only the current session; press d in the picker to set defaults, reducing accidental model drift across sessions.
- MCP tools/list pagination now returns more than the first page. Tool-heavy setups should stop losing capabilities silently.
- Startup hangs when api.anthropic.com is unreachable now time out after 15s, not up to 75s; bad networks hurt less.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.144Read original →
FIG-4121:1
80radar
FIG-4121:1

Sat, May 161 dispatches

#0411
#0411Agents & tools Claude Code Releases4 days ago
`Claude Code` `v2.1.143` adds plugin dependency guards, cost previews, and direct background edits
80radar
Plugin management and background-session controls got noticeably safer and more transparent. The standout changes are dependency-aware enable/disable, projected token cost in the marketplace, and bgIsolation: "none" for repos where worktrees get in the way.
- claude plugin disable now blocks breaking dependency chains and shows a copy-pasteable fix path; enable also force-enables transitive deps.
- The marketplace browse pane now shows per-turn and per-invocation token estimates, making plugin cost visible before turning things on.
- worktree.bgIsolation: "none" lets background sessions edit the main working copy directly, useful when git worktree is too awkward for a repo.
- Background sessions now keep your chosen model and effort after idle wake, fixing a common long-run automation paper cut.
- Worktree cleanup no longer falls back to rm -rf after git worktree remove fails, reducing the chance of losing ignored or in-progress files.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.143Read original →
FIG-4111:1
80radar
FIG-4111:1

Fri, May 151 dispatches

#0410
#0410Agents & tools Claude Code Releases5 days ago
`Claude Code` `v2.1.142`, more controllable background agents and much better daemon reliability
80radar
Background sessions are now configurable from the CLI and far less fragile across sleep, upgrades, worktrees, and MCP timeouts. If you rely on dispatched agents, this is a practical update worth taking now.
- claude agents adds --add-dir, --settings, --mcp-config, --plugin-dir, --model, and --effort. Background runs are easier to pin inside scripts and repeatable workflows.
- Fast mode now defaults to Opus 4.7 instead of 4.6, with CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE=1 as an escape hatch. Speed-oriented usage changes immediately without prompt edits.
- MCP_TOOL_TIMEOUT now correctly affects remote HTTP and SSE MCP servers, removing the old 60-second ceiling. Long-running tool calls are less likely to fail for the wrong reason.
- Sleep/wake reconnects, pre-existing git worktrees, binary-upgrade daemon exits, Chrome-extension crash loops, and Windows network-drive deadlocks were fixed together. Dispatched agents should disappear or wedge far less often.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.142Read original →
FIG-4101:1
80radar
FIG-4101:1

Thu, May 141 dispatches

#0409
#0409Agents & tools Claude Code Releases6 days ago
`Claude Code` `v2.1.141`, hooks and background agents get more usable
60radar
Claude CodeCoding tool — stronger terminal hooks and bg agents
Hook output now supports terminalSequence, and background agents preserve permission mode instead of falling back. This is a solid workflow polish release: fewer paper cuts, better automation hooks, worth updating if Claude Code is in daily use.
- Hooks can now emit terminalSequence JSON for desktop notifications, window titles, and bell events, making local automation flows much easier to wire up.
- claude agents --cwd <path> narrows session lists by directory, which matters once multiple repos and long-running agents pile up.
- Background agents launched with /bg or ←← now keep the current permission mode, removing a subtle safety mismatch in unattended runs.
- /feedback can attach recent sessions from the last 24 hours or 7 days, which improves bug reports for issues that span multiple sessions.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.141Read original →
FIG-4091:1
60radar
FIG-4091:1

Tue, May 121 dispatches

#0408
#0408Agents & tools Claude Code Releaseslast week
`Claude Code` `v2.1.139` adds agent inbox, `/goal`, and plugin cost details
80radar
Claude CodeCoding agent CLI — works toward a goal until done
Claude Code now exposes queued/running/completed agents in one view and can keep working until a goal is met. This pushes it closer to a lightweight async coding orchestrator, so teams already using it should test the new loop controls now.
- claude agents adds a single inbox for running, blocked, and finished sessions; parallel agent work gets easier to monitor without extra tooling.
- /goal now works in interactive, -p, and Remote Control, with live elapsed/turn/token overlay; long tasks need less manual reprompting.
- claude plugin details <name> shows component inventory and projected per-session token cost; plugin overhead becomes easier to budget before rollout.
- Hooks gain args: string[] exec mode and continueOnBlock for PostToolUse; shell-quoting bugs drop and blocked flows can recover automatically.
- Subagent API calls now carry agent lineage headers, and MCP/plugin reload behavior improved; debugging multi-agent setups should get less painful.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.139Read original →
FIG-4081:1
80radar
FIG-4081:1

Sat, May 91 dispatches

#0407
#0407Agents & tools Claude Code Releases2 weeks ago
`Claude Code` `v2.1.136`, fixes MCP auth churn and resume/path edge cases
60radar
Claude CodeCoding agent — AI dev tool for terminal and IDE workflows
Remote MCP setups got more reliable: token refresh races, login loops, and lost sessions were fixed. If your workflow leans on multiple tools and IDE integration, this patch removes daily friction.
- Concurrent MCP token refresh no longer drops refresh tokens, which should stop repeated re-auth across several remote servers.
- /resume and --continue now work when project paths contain underscores, fixing a small but nasty session-recovery edge case.
- IDE integrations no longer lose .mcp.json, plugin, or claude.ai connector servers after /clear, which mattered for long-lived editor sessions.
- settings.autoMode.hard_deny adds unconditional block rules in auto mode, giving tighter guardrails for risky file or tool actions.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.136Read original →
FIG-4071:1
60radar
FIG-4071:1

Fri, May 81 dispatches

#0406
#0406Agents & tools Claude Code Releases2 weeks ago
`Claude Code` `v2.1.133`: worktree base control, effort propagation, multi-session fixes
70radar
Claude CodeCoding tool — an agentic CLI for worktree-based dev
Worktree branching is now explicit, and effort state finally propagates cleanly through hooks and Bash. The bigger win is stability: concurrent sessions, MCP OAuth, subagents, and remote interrupts all got fixes that remove real workflow breakage.
- New worktree.baseRef picks fresh or head, so new worktrees can start from origin/<default> or local HEAD; safer when juggling unpushed changes.
- Hooks now receive effort.level plus $CLAUDE_EFFORT, which makes effort-aware scripts and tool routing much easier to implement.
- A refresh-token race that pushed parallel sessions into 401 failures was fixed; this directly reduces flaky multi-session agent runs.
- MCP OAuth now respects HTTP(S)_PROXY, NO_PROXY, and mTLS across discovery, registration, token exchange, and refresh; enterprise network setups stop randomly breaking.
- Subagents can now discover project, user, and plugin skills again, removing a nasty limitation for delegated workflows.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.133Read original →
FIG-4061:1
70radar
FIG-4061:1

Thu, May 71 dispatches

#0405
#0405Agents & tools Claude Code Releases2 weeks ago
`Claude Code` `v2.1.132` adds session env vars and hardens terminal/resume flows
60radar
Claude CodeCoding agent CLI — stronger terminal recovery and session stability
More of a stability drop than a feature release: terminal recovery, paste handling, resume state, and MCP failure cases all got cleaned up. Worth updating if Claude Code is part of your daily shell or IDE loop.
- CLAUDE_CODE_SESSION_ID now reaches Bash subprocesses, aligning tool env with hook session_id and making per-session logging easier.
- CLAUDE_CODE_DISABLE_ALTERNATE_SCREEN=1 keeps output in native scrollback, useful if you prefer normal terminal history over fullscreen rendering.
- Resume and plan-mode edge cases were fixed, including emoji-corrupted sessions, ignored --permission-mode, and lost plan mode after ExitPlanMode.
- A severe MCP issue was patched: malformed stdio server output could drive 10GB+ RSS memory growth, and failed tools/list now retries with clearer status.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.132Read original →
FIG-4051:1
60radar
FIG-4051:1

Wed, May 61 dispatches

#0404
#0404Agents & tools Claude Code Releases2 weeks ago
`Claude Code` `v2.1.129` adds session plugin fetch and background auto-update
70radar
Claude CodeAI coding tool — terminal-first dev workflow
Plugins are easier to load on the fly, and self-updates now run in the background on package installs. Small release, but the workflow friction drop is real enough to use now.
- --plugin-url <url> pulls a plugin .zip into the current session, which lowers the setup cost for trying custom tooling fast.
- CLAUDE_CODE_PACKAGE_MANAGER_AUTO_UPDATE lets Homebrew and WinGet installs upgrade in the background, then asks for a restart instead of blocking work.
- Gateway model discovery is now opt-in via CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1, reducing surprise model list changes on custom gateways.
- skillOverrides now behaves correctly across off, user-invocable-only, and name-only, which makes shared agent setups easier to control.
- Several regressions were fixed around /clear, subagent UI, prompt cache TTL, and MCP allow rules, so day-to-day CLI reliability improved.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.129Read original →
FIG-4041:1
70radar
FIG-4041:1

`Claude Code` `v2.1.144` improves background sessions, MCP tools, and terminal stability

`Claude Code` `v2.1.143` adds plugin dependency guards, cost previews, and direct background edits

`Claude Code` `v2.1.142`, more controllable background agents and much better daemon reliability

`Claude Code` `v2.1.141`, hooks and background agents get more usable

`Claude Code` `v2.1.139` adds agent inbox, `/goal`, and plugin cost details

`Claude Code` `v2.1.136`, fixes MCP auth churn and resume/path edge cases

`Claude Code` `v2.1.133`: worktree base control, effort propagation, multi-session fixes

`Claude Code` `v2.1.132` adds session env vars and hardens terminal/resume flows

`Claude Code` `v2.1.129` adds session plugin fetch and background auto-update