html-anything
CLAUDE.md10k starsChanged 3 days agoApache-2.0View on GitHub
Read and follow AGENTS.md before working in this repository. It is the shared source of instructions for all Agents. Use CLAUDE.md as the Claude Code instruction entry. Record future repository-rule updates here and keep them aligned with AGENTS.md. Branch new work from the latest main and target all pull requests at main. Merge reviewed changes directly into main when authorized; do not use feature/motion as an intermediate integration branch. Follow the mandatory shared design system rules for every style-related change.…
Trending
More talk lately.thin dataabout the whole repository
Who's talking about nexu-io/html-anything
- A Hacker News story“HTML-anything – The agentic HTML editor – your local AI agent writes the HTML”Hacker News · May 20, 2026 · 3 points
- @anotherbug.com“Markdown 只是草稿 给人看的成品应该直接是 HTML HTML Anything 干的就是这步 笔记、CSV、Excel、JSON 丢进去 它调用你电脑上已经登录的 coding agent …”Bluesky · Sep 2, 2026 · 1 likes
- @agentpalisade.bsky.social“HTML Anything ships no model and no API key. It detects a coding-agent CLI you already logged into and …”Bluesky · Aug 2, 2026 · 1 likes
- nexu-io/open-design“chore(featured): curate Featured picks down to top 10 across categories”GitHub · May 17, 2026 · 1 reactions
- akhmadilham/pinax-registry“[Skill Proposal]: html-anything”GitHub · Nov 3, 2026
The biggest 5 of 13 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 13
Talked about on GitHub, Bluesky and Hacker News (13 times).
On GitHub · 10
- May 17, 2026chore(featured): curate Featured picks down to top 10 across categories — nexu-io/open-design1 reaction · 1 comment
- Nov 3, 2026[Skill Proposal]: html-anything — akhmadilham/pinax-registry0 reactions · 1 comment
- Oct 14, 2026agent-skills.github.nexu-io.html-anything: 0-unstable-2026-09-15 → 0-unstable-2026-10-13 — sudosubin/agents.nix0 reactions
- Oct 13, 2026Embed html-anything skill as Morph AI's base skill for HTML output — FloatLikeABee/robo0 reactions
- Oct 3, 2026agent-skills.github.nexu-io.html-anything: init at 0-unstable-2026-09-15 — sudosubin/agents.nix0 reactions
5 more on GitHub
- Sep 30, 2026[Candidate] nexu-io/html-anything - large collection (81 SKILL.md files) — ProSkillsMD/proskills0 reactions · 1 comment
- Aug 5, 2026idea: nexu-io/html-anything — zqf1314/idea0 reactions · 1 comment
- May 18, 2026Audit candidate: nexu-io/html-anything — xiaolai/nlpm0 reactions · 5 comments
- May 15, 2026docs: add HTML Anything visual page — 53able/docs-site0 reactions
- May 15, 2026TrendForge Daily — 2026-05-15 — anirudhDVDEBUG/RnD0 reactions · 1 comment
On Bluesky · 2
- Sep 2, 2026Markdown 只是草稿 给人看的成品应该直接是 HTML HTML Anything 干的就是这步 笔记、CSV、Excel、JSON 丢进去 它调用你电脑上已经登录的 coding agent 吐一份能发布的单文件 HTML 不用再填一套 API Key Claude… — @anotherbug.com1 likes and repost
- Aug 2, 2026HTML Anything ships no model and no API key. It detects a coding-agent CLI you already logged into and drives it to turn notes or Markdown… — @agentpalisade.bsky.social1 likes and repost
On Hacker News · 1
- May 20, 2026HTML-anything – The agentic HTML editor – your local AI agent writes the HTML3 points
In this repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- add-byok-providerSkill
- openmotion-directorSkill
- openmotion-productionSkill
- openmotion-recreateSkill
- openmotion-reference-videoSkill
- openmotion-template-to-videoSkill
- openmotion-video-to-templateSkill
The file
We split this CLAUDE.md into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 3
Read and follow AGENTS.md before working in this repository.
- Distinctive: line 5
Use CLAUDE.md as the Claude Code instruction entry.
- Distinctive: line 34
Follow MCP onboarding and capability maintenance.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Claude Code
Read and follow AGENTS.md before working in this repository. It is the shared source of instructions for all Agents.
Use CLAUDE.md as the Claude Code instruction entry. Record future repository-rule updates here and keep them aligned with AGENTS.md.
Branch workflow
Branch new work from the latest main and target all pull requests at main. Merge reviewed changes directly into main when authorized; do not use feature/motion as an intermediate integration branch.
Interface styling
Follow the mandatory shared design system rules for every style-related change. Read packages/design-system/README.md, reuse its tokens, styles and assets, and extend the package before adding a missing reusable visual role. Feature modules must reference this source rather than maintain independent palettes or copied styles. Apply the same rule when delegating or reviewing work; authored video and domain geometry retain their own styling boundary. For CI scope and local validation, follow Local validation, including the presentation-document allowlist and the full product-content gate.
Candidate and publication controls
Follow the release workflow manual: separate manual candidate-build and npm-publication entries, plus snapshot-only recovery. Select main for all entries; other execution refs are rejected for new release work. Select test (default, DEV templates) or stable (prod templates). Blank test versions use short workflow-counter test.N numbering; stable requires an explicit new MAJOR.MINOR.PATCH. Keep full run/source identities in the records. Use candidate-install.md for local testing and keep results private. The publication operator enters the exact version and starts the manual workflow; no acceptance checkbox or independent approval environment is enabled. Public candidates must use an exact reviewed source SHA integrated into main. Scan tracked source plus the outer npm tgz and inner application; preserve the positive distribution list and licenses. Keep candidate, native registry installation, real Agent creation and human acceptance evidence separate.
Publication uses the accepted candidate bytes, never a rebuild. Test publishes under candidate, promotes test after registry propagation, archives and updates the DEV guide, then runs non-blocking Mac/Windows cold installs. Stable publishes publicly under stage, requires both exact prod cold installs to pass, then promotes latest, archives and updates the stable guide last; it preserves test. A stage tag does not make the package private. The existing motion-publish-test.yml filename remains the npm Trusted Publisher identity for both channels. Only separately authorized manual publication/recovery may receive npm OIDC or R2 credentials; consumer jobs receive neither. Archive-only work must not publish npm, move tags or access default entries. The read-only Motion installation check can recheck an exact public test or stable version independently. Never overwrite published versions, put private template payloads/internal acceptance material into public artifacts, or infer merge/publication authorization from green CI. Real prod template, Agent and audiovisual acceptance remain release responsibilities; hosted installation checks do not establish them. Skills marketplace publication, main registration, npm trust, main cutover and repository rename remain separate operations.
UI design system and interaction feedback
Follow the shared UI rules and the design-system interaction contract. Transient operation results use useToast and StudioToastViewport on the floating feedback layer, including inside native dialogs; never insert success or failure notices into the content flow. Reuse semantic light/dark tokens, non-modal dropdowns (immediate quick filters; grouped draft selections with live counts and in-panel Apply), padded rounded hover states, one visible focus treatment, and shared loading/fallback geometry. Verify the actual opened states in the browser without shifting or blocking the working area.
Product documentation
Follow the README writing and acknowledgements rule for every README, alternative draft and translation. Keep the main narrative, headings and showcase copy focused on OpenMotion and what each example demonstrates. Put other-product comparisons, inspiration, source-product labels and upstream acknowledgements only in the final Acknowledgements section; do not use headings such as "From MotionGit". Preserve licenses and provenance, and keep required setup identifiers, supported agent/model names and explicitly requested example brand names accurate. Verify this placement before finishing README edits.
The design-system workbench is an independent package-owned editing surface. Keep it free of Studio and business-module imports; UI edits and Agent edits update the same design-system source. Run npm run design:dev to open it and npm run design:test for its save/conflict/independence checks.
MCP onboarding
Follow MCP onboarding and capability maintenance. Changes to local MCP, media generation, Agent adapters, templates or optional presentation must keep the Studio connection page, capability descriptions, setup instructions and compatibility boundaries current in the same change.
Core capabilities and shared APIs
Follow Core capabilities and shared APIs. Every new or extended core capability must support external MCP calls. Product controls and in-app Agents may compose these capabilities according to developer and user needs, using the same API or MCP operations and shared domain implementation. Preserve the same data, job, version, validation and conflict semantics across entry points; keep public discovery, applicable UI consumers, documentation and adapter verification aligned.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

