Claude Code ModsMake Claude Code your workspace.

Anthropic Releases Haiku 5.5 and Lowers Sonnet Cache Pricing Anthropic has released Claude Haiku 5.5. According to the company, it is for cost-sensitive high-volume workloads. For prompts under 100,000 tokens, per million tokens input is $0.10 and output is $0.50, and Sonnet 5.5 cache reads have also been lowered from $0.20 to $0.10 per million tokens. Source: Anthropic · 10/7 https://www.anthropic.com/claude-haiku-5-5

Created my first Claude Mod, hope you can find it useful It shows the 6 numbers Claude Code doesn't: how much of your 5-hour limit is left, when it resets, how full the context is, what the session's tokens cost, how far along your task is, and how long you've been on this project today. Here's what it does 🧵

An article organizing the changes in Claude Code v2.1.300 from a business perspective. It introduces additions to OpenTelemetry events for cache write retention time items, additions around mods and hooks plus validation of hook inputs, fixes for bugs where 400 errors continue after an MCP server tool definition is rejected and where the most recent message is lost when resuming headless execution, support for Claude Tag in Slack Connect channels (disabled by default, enabled by admins), and more.

If I were to follow two repos first, I'd pick official sources: Anthropic Claude Code Releases／CHANGELOG (https://github.com/anthropics/claude-code/releases); OpenAI Codex repo (https://github.com/openai/codex). When updates are frequent, follow releases first; leave commits for tracking specific behavior changes; for Skills／MCP examples, go back to README, source code, and issues to verify, avoiding treating demos as stable practices.

A Stream Deck plugin that shows per-session Claude Code status (working, waiting for permission, done) with pulsing alerts and keys for 5-hour and weekly quota limits, using `claude agents --json` and an optional hook mod.

ANTHROPIC 🔥: Claude Code Desktop now has a new sessions hub, live session statuses on the nav bar, and more! Users can now start, track, manage, and follow up on multiple sessions at once from the Claude Code main screen. This "Claude Hub" experience has been in testing for a while and has gone through several iterations. Interestingly, OpenAI was planning to release a similar experience called "magic notepad" but abandoned it. If you use Claude Code heavily, this is worth testing!

Claude works quietly. Now you can hear it 🎧 Claude Orchestra is a Claude Code mod: typing as it writes, a click for each tool, a chord when it's done. Run agents in parallel and you get an orchestra 🎻 Free, open source, macOS: https://t.co/WDop9hHkqK @claudeai #ClaudeCode https://t.co/whEj1hPZy5

▶ ♫

A Reddit post linking to a GitHub Claude Code mod that visualizes subagents with an expandable tree, timeline and Agent-row cards.

A 3-minute hands-on tutorial covering what Claude Code mods are, installing and disabling them with /plugin, validating with claude plugin validate, authoring one with Claude, and testing offline with claude plugin test.

A Reddit showcase of Changes, a Claude Code mod that adds a /changes pane showing a per-turn timeline of file edits with per-file diffs and a two-press Revert that restores one file to its pre-turn state, plus a companion Receipts that verifies claimed test runs after the last edit. Both are local-only, MIT-licensed, and installable via /plugin from the candygoogy/notary marketplace.

The useful detail in this Claude Code mods catalogue is the permission map: for each mod, it records what it can read, write, run, or send over the network. That’s the kind of context you want before installing a hook. https://t.co/B7yVuK0rrQ

Claude Code is not supposed to look like this. But it does. The mod is Ruflo (ruvnet/ruflo, 74.7K stars and climbing): an agent swarm console inside Claude Code. A hive mind with a queen that dispatches worker agents, and one loop that covers research, ADRs, SOPs, build, test, validate, secure, benchmark, learn. Comment RUFLO and I’ll DM you the GitHub link. Follow first so I can DM you.