🚀 hermesd 2026.6.15 shipped. A read-only terminal dashboard for Hermes Agent — one live screen for your whole ~/.hermes/: gateway, sessions, tokens, cost, tools, cron, skills, logs, memory, kanban, operations, and curator. No API keys. No network. Zero writes to agent state. When you run an agent for real — a gateway across Telegram/Discord/Slack, cron jobs, dozens of sessions spawning sub-agents, 9 providers, skills, memory files — answering "is it healthy and where are my tokens going?" used to mean 6 commands tailing logs. hermesd reads it all and refreshes automatically. 🆕 New in this release: ▸ Curator panel (13) — the newest memory-curation run: before→after skill counts, archived/pruned/added/consolidated, a per-tool call breakdown, the state-transition trail, model/provider, duration, and the LLM summary. ▸ Sessions → Billing & Context — per session: end reason, billing endpoint (billing_base_url), billing mode, and the model's context-window limit joined from context_length_cache.yaml. ▸ Tokens → By Endpoint Cost Status — spend/tokens aggregated per billing endpoint (finer than per-provider), plus a reconciliation line splitting sessions into unknown / subscription-included / estimated. Costs now show authoritative $ for reported/exact/included and ~$ only when truly estimated. ▸ Gateway health — per-platform connection errors (e.g. discord "failed to reconnect"), active-agent count, restart-requested marker, ⚠ in the compact view. ▸ Operations → Response Store stats from response_store.db; PR-monitor reads now cover every naming family with per-repo dedup. ▸ Logs — extra streams: audit, mcp-stderr, workspace, workspace.error. ▸ Kanban — task workflow fields a parent→child decomposition tree. 🔧 Fixed (the important part): The agent had quietly changed several on-disk shapes and hermesd was reading stale formats — so panels silently blanked. Repaired: credential-pool list-vs-dict in auth.json, desktop build-stamp camelCase keys, retired pr-monitor keys, and cost-status vocabulary. To stop this recurring, I added an opt-in contract test (HERMESD_CONTRACT_TEST=1) that runs the collector against real ~/.hermes and fails the instant a populated field reads blank — catching the next drift before users do. 🛡️ Hardening: - Rich markup-injection escaping on all untrusted ~/.hermes free-text (a [/] in a value could crash the TUI) - symlink path-escape guards on every file read (stays inside the Hermes home) - SQL LIKE wildcard escaping in message search - cache-preservation invariant: on any read error, the last-good data stays on screen — the dashboard never blanks ✅ Quality: 777 tests 1 opt-in contract test, 99% coverage, ruff mypy pip-audit clean, CI matrix on Python 3.11/3.12/3.13, trusted-publish to PyPI. Read-only. Fast. Minimal deps (Rich Pydantic PyYAML). Python 3.11 . @NousResearch ⭐ github.com/mudrii/hermesd 📦 pip install hermesd → run hermesd #AIAgents #LLMOps #DevTools #Python #OpenSource #Observability

sqltoerdiagram.com hit 285 on HN doing one thing: paste DDL, get an ER diagram, in your browser, nothing uploaded. The whole product is a static file on a CDN. The whole moat is the network tab staying empty. Every dbdiagram.io / drawSQL competitor assumed they needed servers and accounts. This one assumed senior devs wouldn't paste prod schemas into someone else's database — and was right. #devtools #security

I got tired of juggling six extensions DevTools just to inspect a page, so I built one that does it all locally I do a lot of web and SEO work, and my browser had turned into a graveyard of single-purpose extensions — one for fonts, one for color picking, an SEO checker, a contrast checker, a tech-stack detector — plus DevTools open on the side. And half of them wanted permissions I didn't love. So I built the one I actually wanted: A1 Multitool. Open the popup, pick a mode, and it inspects the current tab — locally, and only after you click. What's in it: - Type: hover-highlight any element to read its computed font, size, weight, spacing, color; lists every font on the page - Color: sample colors, copy HEX/RGB/HSL, view the palette, check contrast - SEO: indexability, title/description lengths, canonical robots, OG/Twitter tags, JSON-LD, headings - Links: internal vs external, nofollow usage, external domains - HTML: duplicate IDs, empty attributes, missing alt/lang/title - Accessibility: unlabeled inputs, nameless buttons, missing alt, heading skips, landmarks - Performance: load milestones, request count, transfer size, largest resources, third-party hosts - Tech stack: frameworks, analytics, ad tech, consent tools, CDNs - Overlay: outline elements or hover to measure pixel sizes The part I care most about: it runs entirely on-device. No page content, reports, browsing history, or settings get sent anywhere. Every check runs locally, on the current tab, after you click. It's early and I'd genuinely like feedback — what's missing, what's clunky, and what you'd actually reach for day to day. Here's the link chromewebstore.google.com/de…

Your flagship model shouldn't be running grep 😮‍💨 Atlas Cloud now plugs into OpenClaude by @gitlawb, the open-source terminal coding agent, with one key. 🤯Route each role to its best model: top reasoning for planning, fast low-cost models for execution. #API #OpenClaude #AtlasCloud #CodingAgents #DevTools #GenAI #OpenSourceAI #LLMRouting #AICoding

フィードバックありがとうございます 生存時間の優位差はあるのでリリースはしようとは思うのですが、Chromeがメモリをどれくらい使っているかによってDevToolsがGCによって落とされやすいという感じなので、艦これしつつ他作業もしつつだとこれは落ちやすいのかもしれないなと思っているところです

no claude no burp no postman just me and devtools 💞

Running Claude Code, Codex, and Gemini CLI in three terminals at once is genuinely painful. AionUi fixes that by putting 20 coding agents into one free desktop app. You run them side by side, share context between them, and just let them work. It's local-first, works with any API key, and you can check on things from your phone. One window instead of a terminal graveyard. ⭐ 28.2K #AI #DevTools github.com/iOfficeAI/AionUi Follow for daily dev finds 🔔

公式changelogでは、2026-06-11のCodex app 26.609でDeveloper modeが追加され、Chrome DevTools Protocol経由でnetwork/console/runtime errors/page stateを扱えると説明されている。Browser useもCDP/DOM snapshot最適化で最大2倍高速化。これは便利機能というより、agentにブラウザ検証を任せる時の監査境界の話。 developers.openai.com/codex/…

Kage crawls any website and compiles the crawl into a single @golang binary — HTML, CSS, JS, fonts, the whole tree baked into one executable via embed.FS. 214 on HN. The novelty isn't the crawler or the asset rewriting (both 2015 tech). It's that the output is an executable, not a folder of HTML you have to host. Double-click, visit localhost, done. No installer, no static server, no node_modules. The use cases that surfaced in the thread weren't archival nostalgia — they were docs versioning, air-gapped enterprise, compliance snapshots, and conference handouts. WARC is still correct for 50-year horizons. For 'works next Tuesday on a laptop without WiFi,' this is the right shape. #OpenSource #devtools

Some devtools startup called "reload in 5" vibeforked my DC Tycoon Game Somehow theirs is slow af and buggy af though 🤣🤣🤣 But I love that they have 3D visualisation of the racks!

test 2, the network tab (5 min): open devtools, Network tab, refresh, click the requests and read what comes back. data that isn't yours sometimes downloads even when the screen hides it. then Sources tab, search "sk_". a secret key there means anyone loading your site can read it.

Open a PR. Get a live preview. That's it. Preview deployments, make changes, and test out your project before it goes live. Try it today! withnubo.com #DevTools #ShipFast #Deploy

I just checked it myself, and you can as well. Use your computer and go to sooon.fun. Go to Devtools (F12) and network. Look at the calls. It's a centralized server with no blockchain calls that I could find. Prove me wrong.

If you're shipping AI features and tired of API key sprawl: → zyloo.io One key. All models. Less headache. #AI #API #DevTools

agentwire v1.38.0: `agentwire worktree <name>` spins up an isolated branch worktree its own session in one shot. also retired our whole legacy missions dispatcher — turns out primitives a scheduler beat a bespoke PM layer. github.com/dotdevdotdev/agen… #BuildInPublic #DevTools

Chromiumのコード量がLinuxカーネルより多い！」って驚いてる人いるけど、むしろ当然では？ LinuxカーネルはOSの中核部分。ユーザーランドやアプリは含まない。 一方ChromiumはHTML/CSS/JSエンジン、GPUレンダラ、動画再生、WebRTC、PDF、DevToolsなど巨大な機能の集合体。 今のブラウザは実質“Web向けOS”みたいなものだから、コード量がカーネルを超えても不思議じゃない。コード量＝技術的難易度でもないし。

AI 今朝3本 1. Grok Build、ターミナルにLaTeX/数式を表示。ツール往復が不要に。 2. Fable 5/Mythos 5、米輸出規制で全停止（6/12）。復旧交渉中。 3. ChatGPT、記憶の個別編集・削除に対応。CodexにブラウザDevTools検査。 今週はFable復活するかな？GPTの新モデルくるか？Grokも強くなるか？？

Aldform gives devs full form infrastructure - build your own HTML/CSS, tag fields with one attribute, use two endpoints and a tiny SDK to capture submissions, route them and run custom post-submit logic. #WebDev #Forms #DevTools

[ Chromium ] Stealth request that bypasses CSP, hides from DevTools, and leaks the real User-Agent (in case you faked it 😁) brokenbrowser.com/blog/2026-…