Skip to content
DevAgentRadarDevAgentRadar
Sign in

Personal · Free · No per-user AI cost

My week

What mattered for the agents you use. Industry judgment is shared; My stack slices it for you — no login required, optional Google sync. RSS of this week is the same judged list as a feed — no account, no per-subscriber AI.

View filter:

Industry releases

5

Open proposals

3

Tools watched

all

Config that moved

Stated contract changes across the judged radar — config you own, not a vendor leaderboard. Mentions without a change are omitted. How a harness is layered

hooks1 change · latest Sep 3, 2026 · Claude Code v2.1.260 ↗

Claude Code

permission rules1 change · latest Sep 3, 2026 · Claude Code v2.1.260 ↗

Claude Code

MCP server config1 change · latest Aug 27, 2026 · Goose v1.48.0 ↗

Goose

agent context files1 change · latest Sep 3, 2026 · Claude Code v2.1.260 ↗

Claude Code

context and compaction settings1 change · latest Sep 3, 2026 · Claude Code v2.1.260 ↗

Claude Code

background and headless runs1 change · latest Sep 9, 2026 · Google Antigravity 1.1.28 ↗

Google Antigravity

skills and plugins1 change · latest Sep 3, 2026 · OpenAI Codex CLI rust-v0.153.0 ↗

OpenAI Codex CLI

sandbox settings1 change · latest Sep 3, 2026 · Claude Code v2.1.260 ↗

Claude Code

Shared industry brief

Radar brief — 2026-09-09

CLI coding agents spent this window on permission correctness, MCP/provider plumbing, and headless reliability — while Windsurf removed Cascade from Devin Desktop and GPT-5.6 routing showed up in multiple GitHub changelogs.

Read brief

Decide these proposals

All proposals →
proposalteamP5Worth a lookCline

Move Agent Plugins into ~/.agents/plugins and Hub-enable them in Settings → Customize

Cline Desktop v0.0.23 discovers Agent Plugins only from ~/.agents/plugins via the shared Hub. Workspace .agents/plugins is ignored. Put packages with plugin.json there so skills and MCP servers start automatically, then enable each plugin under Settings → Customize.
  1. mkdir -p ~/.agents/plugins && if [ -d .agents/plugins ]; then echo 'Workspace .agents/plugins is ignored in desktop-v0.0.23 — copy each package into ~/.agents/plugins/<name> with its plugin.json'; ls -la .agents/plugins; fi
  2. For every directory in ~/.agents/plugins, open plugin.json and keep only packages whose declared Agent Skills and MCP servers (stdio, Streamable HTTP, or SSE) you want the Hub to validate and start automatically.
  3. In Cline Desktop open Settings → Customize, use the Agent Plugins list (not Cline Plugins), read each plugin's description, badge, and contributed tools, then Hub-enable only the plugins this machine should run.
agentmcpsecuritycliRelease source ↗

Your loop

This browser · no sign-in · not shared as “you”

After you run the prompt

Only you can mark this. Agents cannot write your loop.

Your decision stays on this device. A public tally appears after a few votes.

proposaleveryoneP4Worth a lookCline

Pin Cline models after the v3.0.61 switch to Claude Fable 5.1

Cline CLI v3.0.61 changes the resolved default for 57 providers (Anthropic and many gateways now Claude Fable 5.1 instead of Opus 5). Upgrade, pin every model you intend to keep, and adopt Hub-drain, 10s MCP connect, and ~/Cline/Rules so sessions no longer stall or silently change.
  1. Upgrade Cline to CLI v3.0.61. On next launch, if prompted that the Hub is older than the CLI, press Enter to replace it so in-flight turns drain first, or Escape to keep the current Hub; do not kill a Hub that is too old or wedged to accept drain.
  2. Pin a model on every Cline provider you use. After v3.0.61, unpinned Anthropic, Amazon Bedrock, Vertex, OpenRouter, Vercel AI Gateway, Kilo Gateway, LLM Gateway, DevPass, DigitalOcean, CrossModel, Eden AI, and NanoGPT resolve to Claude Fable 5.1 instead of Claude Opus 5 (57 providers change defaults).
  3. For Dify, SAP AI Core, opencode, and Codex CLI, restore tool-calling and image file-read workflows: v3.0.61 no longer treats an empty catalog capability list as an authoritative denial that strips every tool and drops images.
agentmcpmodelpricingcliRelease source ↗

Your loop

This browser · no sign-in · not shared as “you”

After you run the prompt

Only you can mark this. Agents cannot write your loop.

Your decision stays on this device. A public tally appears after a few votes.

proposaldeveloperP4Worth a lookxAI Grok CLI (Grok Build)

Replace <output_efficiency> with <response_guidelines> in Grok CLI prompts

Grok CLI renames the system prompt tag <output_efficiency> to <response_guidelines>, strips trailing /** on sandbox allow paths, restricts session dirs to owner-only, and installs native arm64 on Apple Silicon (including Rosetta/x86_64 updaters).
  1. Search the repo, prompt overlays, snapshots, and string assertions for the literal tag <output_efficiency> and rename it to <response_guidelines> so Grok CLI be71313 / Source-Revision 5d08d7e4123092567ccd584cd9f99afa2972065c still applies response guidelines.
  2. For each Grok CLI sandbox allow path, strip a trailing /** (the CLI now does this) and re-test that the remaining prefix still allows only the intended directories and denies writes/reads outside it.
  3. Update every mkdir/umask (and chmod existing session dirs) so Grok CLI session directories are owner-only (0700), matching restrict-session-directories-to-owner-only-permissions.
agentmodelsecurityclibreakingRelease source ↗

Your loop

This browser · no sign-in · not shared as “you”

After you run the prompt

Only you can mark this. Agents cannot write your loop.

Your decision stays on this device. A public tally appears after a few votes.

High-signal releases

Goosev1.48.0High impactAug 27, 2026

TrustedRouter declarative provider [#11422](https://github.com/aaif-goose/goose/pull/11422)

  • OpenCode Zen gateway as a declarative provider [#10939](https://github.com/aaif-goose/goose/pull/10939)
  • Gondola as declarative OpenAI-compatible provider [#11421](https://github.com/aaif-goose/goose/pull/11421)
  • SayGM as declarative OpenAI-compatible provider [#11267](https://github.com/aaif-goose/goose/pull/11267)
  • Lynkr as a declarative OpenAI-compatible provider [#11372](https://github.com/aaif-goose/goose/pull/11372)
  • +107 more changes
Read the full release note

✨ Features

Providers

  • TrustedRouter declarative provider #11422

  • OpenCode Zen gateway as a declarative provider #10939

  • Gondola as declarative OpenAI-compatible provider #11421

  • SayGM as declarative OpenAI-compatible provider #11267

  • Lynkr as a declarative OpenAI-compatible provider #11372

  • PleumRouter declarative provider #10479

  • Z.ai GLM-5.2 replaced with GLM-5.3 #11226

  • Custom provider cost fields drive cost tracking #11220

  • Model-native audio transcription provider #10589

  • OpenAI GPT-5.6 (sol/terra/luna) routing via AWS Bedrock #10502

  • GPT-5.6 follow-up support for Codex and Responses API #10460

Hooks

Excerpt ends here — this release continues at the source ↗.

Source ↗
Claude Codev2.1.260High impactSep 3, 2026

Added a diff panel that opens beside the conversation in fullscreen mode and shows your uncommitted changes as Claude edits; toggle it with `/diff`

  • Added a likely cause for prompt-cache misses (e.g. tool definitions or system prompt changed, idle past the TTL) to `/cost` and the status line's `prompt_cache` field
  • Added `/reload-plugins` to headless sessions, so it appears in the Claude Code Desktop and SDK command lists
  • Added a text form of `/advisor` (`/advisor`, `/advisor <model>`, `/advisor off`) for the desktop app, Remote Control, and other headless (`-p`/Agent SDK) sessions
  • +71 more changes
Read the full release note

What's changed

  • Added a diff panel that opens beside the conversation in fullscreen mode and shows your uncommitted changes as Claude edits; toggle it with /diff
  • Added a likely cause for prompt-cache misses (e.g. tool definitions or system prompt changed, idle past the TTL) to /cost and the status line's prompt_cache field
  • Added /reload-plugins to headless sessions, so it appears in the Claude Code Desktop and SDK command lists
  • Added a text form of /advisor (/advisor, /advisor <model>, /advisor off) for the desktop app, Remote Control, and other headless (-p/Agent SDK) sessions
  • Added oidc.scope_on_refresh to the Claude apps gateway for IdPs that return an id_token on refresh only when asked for openid again
  • Added Claude apps gateway support for newer Claude Desktop keys in desktop policy blocks, including userPluginMarketplacesEnabled and userPluginUploadsEnabled
  • Fixed Edit/Write/Read permission rules whose path contains parentheses being dropped as invalid or ignored by the Bash sandbox, which left "read-only" folders writable

Excerpt ends here — this release continues at the source ↗.

Source ↗
Zed AIv1.19.0-preHigh impactSep 2, 2026

This week's release includes call hierarchy support, multi-select in the Git Panel, automatic language detection for untitled buffers, and project search on type by default.

  • Added automatic language detection for untitled buffers. ([#61412](https://github.com/zed-industries/zed/pull/61412); thanks [amtoaer](https://github.com/amtoaer))
  • Added support for rendering LSP 3.18 Markdown diagnostic messages from language servers. ([#61030](https://github.com/zed-industries/zed/pull/61030); thanks [aviatesk](https://github.com/aviatesk))
  • +58 more changes
Read the full release note

This week's release includes call hierarchy support, multi-select in the Git Panel, automatic language detection for untitled buffers, and project search on type by default.

Shipped by the Zed Guild 🛡️

  • Added automatic language detection for untitled buffers. (#61412; thanks amtoaer)

  • Added support for rendering LSP 3.18 Markdown diagnostic messages from language servers. (#61030; thanks aviatesk)

  • Fixed searching for full SHA-256 commit hashes in the Git Graph. (#63371; thanks albertbogusz)

Learn about the Zed Guild.

Features

AI

  • Added support for effort-based reasoning when using OpenRouter models. (#61308; thanks codello)

  • Added support for renaming Terminal Threads from the Threads Sidebar. (#63494)

Excerpt ends here — this release continues at the source ↗.

Source ↗
Google Antigravity1.1.28High impactSep 9, 2026

Improved resilience to transient model API errors: the agent now retries errors such as `503 Unavailable` for much longer with exponential backoff, so a brief service hiccup no longer aborts your session.

  • Improved sign-in and startup speed by reading your signed-in identity from the stored credential instead of making a network request on every launch and after browser consent; a failed identity lookup no longer prevents sign-in from completing, and the displayed plan tier no longer briefly disappears during startup.
  • +14 more changes
Read the full release note
  • Improved resilience to transient model API errors: the agent now retries errors such as 503 Unavailable for much longer with exponential backoff, so a brief service hiccup no longer aborts your session.
  • Improved sign-in and startup speed by reading your signed-in identity from the stored credential instead of making a network request on every launch and after browser consent; a failed identity lookup no longer prevents sign-in from completing, and the displayed plan tier no longer briefly disappears during startup.
  • Improved headless (-p) runs to exit promptly once the final answer is delivered: the CLI waits for running background tasks and scheduled timers to finish, bound by --print-timeout, and leaves daemon background tasks such as dev servers running instead of terminating them.
  • Improved failure reporting in headless (-p) runs: fatal errors now appear on stderr with a stable error: marker, a note is printed when the response may be truncated, and runs that previously ended silently with no output now explain why.

Excerpt ends here — this release continues at the source ↗.

Source ↗
OpenAI Codex CLIrust-v0.153.0High impactSep 3, 2026

Vim mode now supports undo with `u` and redo with `Ctrl+R`, preserving complete drafts including pasted content and attachments. ([#41941](https://github.com/openai/codex/pull/41941), [#42140](https://github.com/openai/codex/pull/42140))

  • The plugin CLI can list, install, and remove plugins from remote marketplaces. ([#42150](https://github.com/openai/codex/pull/42150))
  • Set `tui.auto_recap = false` to disable automatic recaps while keeping manual `/recap` available. ([#42101](https://github.com/openai/codex/pull/42101))
  • +99 more changes
Read the full release note

New Features

  • Vim mode now supports undo with u and redo with Ctrl+R, preserving complete drafts including pasted content and attachments. (#41941, #42140)

  • The plugin CLI can list, install, and remove plugins from remote marketplaces. (#42150)

  • Set tui.auto_recap = false to disable automatic recaps while keeping manual /recap available. (#42101)

  • TUI history shows complete patches, input sent to background terminals, and individual completed commands. (#41893, #42107)

  • Plus and Team users receive an earlier warning when less than half of their allowance remains in an approximately five-hour usage window. (#42142)

Bug Fixes

  • TUI sessions reconnect after an external app-server connection drops, preserving drafts and transcripts while keeping uncertain or queued submissions paused for review.

Excerpt ends here — this release continues at the source ↗.

Source ↗