mudrii

OpenClaw Dashboard v2026.4.13 is out. This release ships the diagnostics work requested in issue #14 and includes a full hardening pass before release: - better logs/errors visibility - tighter refresh + shutdown lifecycle - correct gateway timeout handling - improved merged log selection - fixed subagent accounting - updated docs + release packaging Also includes merged improvements from PRs #19 #20 #21 #22 by @SweetSophia. Release: github.com/mudrii/opencla… #golang #opensource #openclaw

Just released hermesd — a TUI dashboard for monitoring Hermes Agent (@NousResearch's self-improving AI agent). One command. Eight panels. Real-time visibility into sessions, tokens, costs, tools, cron, skills, and logs. Read-only. Zero config. Works over SSH. pip install hermesd github.com/mudrii/hermesd 🧵 What it shows: [1] Gateway — PID, version, platform connections [2] Sessions — active count, messages, tool calls [3] Tokens — today's usage + estimated cost [4] Tools — available tools + call stats [5] Config — model, provider, personality [6] Cron — jobs, schedule, errors [7] Skills — 69 skills by category with descriptions [8] Logs — agent/gateway/errors with tab switching Why I built it: Running Hermes Agent with Telegram + Discord + cron + multiple models = information scattered across 5+ sources. hermesd reads ~/.hermes/ and shows everything in one live-updating terminal dashboard. Never writes to your agent state. 168 tests. 3 dependencies (all already in Hermes Agent). Install: pip install hermesd # or uv tool install hermesd Then just: hermesd Works on macOS and Linux. SSH/tmux compatible. Adapts layout to your terminal size. MIT licensed. PRs welcome. pypi.org/project/hermes…

OpenClaw Dashboard v2026.4.8 is live. This release rolls up the biggest improvements from the last few releases into one update: - built-in service commands: install/start/stop/restart/status - macOS launchd + Linux systemd --user support - fixed release packaging + direct-download archives - stronger Homebrew install/upgrade behavior - better service health checks on install - fixed background-service env/path issues for openclaw - safer routing/CORS/error-path handling - more tests + overall hardening Net result: easier installs, cleaner upgrades, more reliable background operation. github.com/mudrii/opencla… #OpenClaw #golang #opensource #devtools

gobird 26.04.08 is out. This latest release includes everything shipped across the last 3 releases: - public OSS release of the X/Twitter CLI + Go client - read/search/post/reply/bookmarks/timelines/lists/news/trends - Safari/Chrome/Firefox auth - Homebrew install - stronger auth + cookie validation - safer rate limiting, pagination, media upload, and query ID refresh - concurrency-safe throttling with cancelation support - cleaned config surface + synced docs - upgraded to Go 1.26 Repo: github.com/mudrii/gobird Install: brew install mudrii/tap/gobird #golang #opensource #cli

Claude Code oauth giving me hell for the 3rd time in 2 wks fml

is Claude Code OAuth busted right now?

OAuth error on Claude Code. github.com/anthropics/cla…

is @AnthropicAI oAuth for Claude Code down, haven't been able to authenticate for the past two hours? anyone else facing this issue?

@sytaylor @AnthropicAI Yeah, my OAuth is broken, too. Just for regular claude code purposes

Claude CodeのOAuthが落ちているのか仕事にならん・・・

@AnthropicAI Claude Code Oauth Down？

@theo Anthropic’s subscription oauth with claude code is broken today, too. Literally unusable

Claude Codeの /loginがずっとOAuth error: timeout of 15000ms exceededになって一生解決出来ない。

error in my claude code cli "OAuth error: timeout of 15000ms exceeded" i cant login

i can't believe more people aren't talking about this part of the claude code leak there's a hidden feature in the source code called KAIROS, and it basically shows you anthropic's endgame KAIROS is an always-on, *proactive* Claude that does things without you asking it to. it runs in the background 24/7 while you work (or sleep) anthropic hasn't turned it on to the public yet, but the code is fully built here's how it works: every few seconds, KAIROS gets a heartbeat. basically a prompt that says "anything worth doing right now?" it looks at what's happening and makes a call: do something, or stay quiet if it acts, it can fix errors in your code, respond to messages, update files, run tasks... basically anything claude code can already do, just without you telling it to but here's what makes KAIROS different from regular claude code: it has (at least) 3 exclusive tools that regular claude code doesn't get: 1. push notifications, so it can reach you on your phone or desktop even when you're not in the terminal 2. file delivery, so it can send you things it created without you asking for them 3. pull request subscriptions, so it can watch your github and react to code changes on its own regular claude code can only talk to you when you talk to it. KAIROS can tap you on the shoulder and it keeps daily logs of everything. > what it noticed > what it decided > what it did append-only, meaning it can't erase its own history (you can read everything) at night it runs something the code literally calls "autoDream." where it consolidates what it learned during the day and reorganizes its memory while you sleep and it persists across sessions. close your laptop friday, open it monday, it's been working the whole time think about what this means in practice: > you're asleep and your website goes down. KAIROS detects it, restarts the server, and sends you a notification. by the time you see it, it's already back up > you get a customer complaint email at 2am. KAIROS reads it, sends the reply, and logs what it did. you wake up and it's already resolved > your stripe subscription page has a typo that's been live for 3 days. KAIROS spots it, fixes it, and logs the change endless use-cases, it's essentially a co-founder who never sleeps the codebase has this fully built and gated behind internal feature flags called PROACTIVE and KAIROS i think this is probably the clearest signal yet for where all ai tools are going. we are heading into the "post-prompting" era where the ai just works for you in the background like an all-knowing teammate who notices and handles everything, before you even think to ask

libghostty for Rust! Some Ghostty maintainers came together and made a really high quality Rust crate in front of libghostty-vt. The repo includes a full port of Ghostling (full terminal in a single Rust file!). The API is beautiful. github.com/Uzaaft/libghos…

Shipped a new release of gobird (v26.03.24) — an open-source CLI tool and Go client library for Twitter/X. This release focused on security hardening and code quality: Security - Credential validation on the public API to prevent HTTP header injection - Path traversal protection on Chrome cookie extraction - CI actions pinned to immutable commit SHAs Bug Fixes - Rate limiter no longer blocks concurrent callers during sleep - Server-controlled Retry-After values capped to prevent indefinite hangs - CLI flag shadowing fixed across 10 subcommands - Deterministic media variant selection - Thread parsing now correctly handles newest-first tweet ordering Quality - 20 new tests covering every fix - 0 lint issues, 0 race conditions - golangci-lint clean (errcheck, gosec, govet, staticcheck, revive) Now installable via Homebrew: brew install mudrii/tap/gobird GitHub: github.com/mudrii/gobird #golang #opensource #cli #security #twitter

The stack: → Layer 1: Flat files for boot context (SOUL.md, MEMORY.md) → Layer 2: Semantic search over workspace + vault (QMD + sqlite-vec) → Layer 3: Auto-capture from conversations (Mem0 + Ollama) → Layer 4: Curated knowledge graph (Obsidian vault) Each layer solves a different memory problem. Post 3 (reply thread): Key insight: 4 agents writing to the same knowledge base = chaos. Solution: zone ownership. Each agent writes to its own zone, searches everything, and promotions go through an inbox pipeline with provenance tracking. #AIAgents #LLM #MemoryArchitecture #AgenticAI #BuildInPublic #OpenSource

I built a 4-layer memory system for my AI agent setup and wrote about what actually worked (and what didn't). The hard part isn't storing information — it's knowing what to remember, when to recall it, and how to keep it fresh. x.com/mudrii/status/… Original Blog Post: blog.mudreac.com/post?slug=mult… (blog.mudreac.com/post?slug=mult…)