* fix(mcp): eager-auth flag on the Grok connector links so Grok starts OAuth Live test after #2158: pasting the Grok URL into grok.com's custom connector dialog listed all 150+ tools and never opened the sign-in. Grok probes the URL without credentials, like claude.ai, and reads the lazy 200 on initialize as an authless server; only the 401 challenge starts OAuth (#2159 fixed the same thing for the claude.ai link). - lib/onboarding/checklist.ts: mcpServerUrl() builds the server URL with an optional eagerAuth flag; sideDoorServerUrl() gives the Grok side door auth=required and keeps ChatGPT lazy; claudeConnectorLink() reuses it. SIDE_DOORS / SideDoor move here from the component. Tests for all three. - NewUserChecklist copies the door-specific URL (now with a client marker). - ApiKeysPanel's Grok row copies the flagged URL, mirroring the Claude one. - auth-mode.ts comment records the second consumer; registry entry's Grok step carries the flag; DECISIONS. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LhTJcwgzmN3TsLR8tVHwdi Signed-off-by: Emil <emilmattsson14@gmail.com> * docs(mcp): registry Claude.ai step carries auth=required too Review pass on #2167: the registry entry flagged the Grok install URL but left the Claude.ai step on the bare URL, which pre-fills "None" in claude.ai's dialog (#2159). Same file, same flag, now consistent. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LhTJcwgzmN3TsLR8tVHwdi Signed-off-by: Emil <emilmattsson14@gmail.com> --------- Signed-off-by: Emil <emilmattsson14@gmail.com> Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Community Registry
This directory is the source of truth for the registry shown at gnubok.se/community/registry: skills, MCP servers, workflows and apps built on Accounted. The website syncs its registry pages from here, so adding an entry is a normal pull request against this repo.
Structure
registry/
entries/ one .mdx file per registry entry
authors/ one .mdx file per author profile
Files starting with _ are templates and are ignored by validation and sync.
Adding an entry
- Copy
entries/_template.mdxtoentries/<your-slug>.mdx. The filename is the slug and the public URL:gnubok.se/community/registry/<your-slug>. - If this is your first contribution, copy
authors/_template.mdxtoauthors/<your-handle>.mdx. Theauthorfield in your entry must match an author file, so first-time contributors add both files in the same PR. - Validate locally:
npm run validate:registry(ornpx tsx scripts/validate-registry.tswithout installing everything). - Open a PR. Commits need a DCO sign-off (
git commit -s), same as the rest of the repo; see CONTRIBUTING.md.
A maintainer reviews the entry (does it work, does it describe itself honestly, is the content safe) and merges. After merge the website pulls the entry in with its registry sync; expect it live on the site within a few days of merging.
Entry frontmatter
Required:
| Field | Type | Notes |
|---|---|---|
title |
string | Shown as the card and page title |
description |
string | Card subtitle, max 500 chars; put detail in the body |
slug |
string | Must equal the filename |
kind |
skill | mcp | workflow | app |
Which shelf it goes on |
author |
string | Handle of a file in authors/ |
status |
live | beta | archived |
Be honest; beta is fine |
lang |
sv | en |
Language of the entry body |
personas |
list | Any of founder, finance, byra, developer |
publishedAt |
date | YYYY-MM-DD |
updatedAt |
date | YYYY-MM-DD, bump when you edit |
At least one of installCommand, downloadUrl, repoUrl, externalUrl is
required so readers can actually get the thing. Optional extras:
featured (maintainer-set), oauthScopes, gnubokTools (which MCP tools it
touches), requiresWriteScope, sieCompatible, version, faq (list of
{q, a}), related (list of slugs), ogImageEyebrow.
Body rules
The body after the frontmatter is plain Markdown (headings, lists, tables,
links, fenced code blocks). The website renders bodies through MDX, which
evaluates raw tags (lowercase HTML included), {...} expressions and
import/export statements at build time, so the validator rejects all of
them. Fenced code blocks and backtick inline code are fine; they are
displayed, never executed. Need a literal <tag> or {value} in prose? Put
it in backticks.
Write the body like documentation, not a landing page: what it does, what it needs (scopes, API keys), what it will not do, and one honest limitation beats three superlatives.
What gets accepted
- It must exist and work against Accounted today (the live app or the MCP server).
- It must describe its write behavior truthfully. Anything that writes to the ledger goes through staged operations that a human approves; entries that work around that will not be listed.
- No fees are charged for listing, and your IP stays yours. Entries are documentation and are contributed under this repo's license; the thing the entry points to keeps whatever license you gave it.