The branding definitely doesn't help. I'd argue that a simple website landing page and just keeping the old forum style stuff in a readonly state would just be a good first step.
The server had three env vars to restrict which Kubernetes tools agents could use: readonly-only, non-destructive-only, and a custom allowlist. Operators configured them, trusted them, deployed to production.
Windows just got a new driver. 🤖
@AnthropicAI's @claudeai is no longer vision-only.
Screenshot vision — up to 26s a turn — is now the FALLBACK. The default: Claude reads the Windows UI tree in ~20ms and drives it cursor-free, in the background.
• Clicks & types the live Chromium / Electron / VS Code web DOM
• OCR reads raw pixels (games, canvas) → clicks the text it sees
• Even occluded & GPU-composited windows
Built on @bunjavascript — zero native deps, no build step, one line:
claude mcp add uia -- bunx bun-uia
46 policy-gated tools · readonly / safe / full
It reads the UI Automation tree, not pixels — finds controls by name/role, acts cursor-free, and sees occluded/background windows via Windows.Graphics.Capture.
36 policy-gated MCP tools, deployer picks readonly / safe / full. 🧵
To write a comment you wrap it in special characters <!-- and -->
And you do it like so
<button readonly> this button is readonly, you cannot click it, it will not work & yes the opening tag is weird kindly see comment</button>
----
New QStash feature: Readonly tokens! 🎉
◆ Give read-only access to your message queue
◆ Can't send, modify, or delete messages
Live on the QStash dashboard 👇
Openclaw ist wie ich finde granularer, bastellastiger, aber dadurch auch extrem mächtig. Hermes ist mehr straight forward out of the box.
Habe für Hermes auch einen eigenen User und eigene Accounts wie z. B. eine GMX-Mailadresse, aber mit (zunächst) Readonly-Freigabe auf mein persönliches Postfach.
Von Goose hab ich bisher noch gar nix gehört, muss ich mal reinschauen
Readonly properties in #PHP 8.1 don't use a runtime check on every read.
The engine sets a ZEND_ACC_READONLY flag on the property at class link time. Reads go through the normal path, no extra cost.
Only WRITE opcodes (ASSIGN_OBJ and friends) check the flag. And only on the second write, since the first one is the initialization.
Free reads, throw on rewrite. That's the entire mechanism!
in 4.2, i will add proper audit policies, members will be readonly, admins and owners will have full access. it is a good core for adding more related stuffs, like user assignments for each resource
Claude Code CLI 2.1.176 changelog:
New feature:
• Added footerLinksRegexes setting for regex-matched link badges in the footer row, configurable via user or managed settings
Fixes:
• Fixed availableModels enforcement: alias model picks can no longer be redirected to a blocked model via ANTHROPIC_DEFAULT_*_MODEL environment variables, and /fast now refuses to toggle when it would switch to a model outside the allowlist
• Fixed auto mode failing on Fable 5 for organizations without Opus 4.8 enabled — the classifier now falls back to the best available Opus model
• Fixed hook if conditions for Read/Edit/Write tool paths: documented patterns like Edit(src/), Read(~/.ssh/), and Read(.env) now match correctly
• Fixed Linux sandbox failing to start when .claude/settings.json is a symlink with an absolute target
• Fixed /copy and mouse-selection copy not reaching the system clipboard inside tmux over SSH, and tmux paste buffer not loading on versions older than 3.2
• Fixed Remote Control connecting from web/mobile silently switching the session's model
• Fixed Remote Control disconnect notifications showing a bare numeric code instead of a human-readable reason, and connection failures adding a duplicate line to the conversation transcript
• Fixed Remote Control sessions not disconnecting when you sign in to a different account
• Fixed /cd and worktree moves leaving the session reporting the previous directory's git branch
• Fixed claude agents: pressing back in one window no longer detaches other windows attached to the same session
• Fixed backgrounded sessions showing "Working" forever when /bg mid-turn had nothing left to continue
• Fixed background agent search by PR URL: PRs opened during scheduled wakeups or while a job was blocked now appear in claude agents search
• Fixed the agents view input showing no text cursor on Windows
• Fixed claude --bg -cn <name> not seeding the session name
• Fixed background sessions to neutralize Windows network paths in persisted state before respawn
• Fixed background-session respawn rejecting malformed resume IDs from corrupted state files
• Fixed the Windows background-service daemon not starting when ~/.claude/daemon has the ReadOnly attribute set
• Fixed cloud sessions failing with "Could not resolve authentication method" when idle for too long before being claimed
Improvement:
• Improved Bedrock credential caching: credentials from awsCredentialExport are now cached until their Expiration instead of a fixed 1 hour
Other changes:
• Session titles are now generated in the language of your conversation (set the language setting to pin a specific language)
• Background sessions now show clearer guidance when a window left open across an auto-update can't submit a reply, and claude daemon status explains version-skew behavior
Source: github.com/anthropics/claude…