Use Vincia in your AI client
Vincia exposes a public MCP server at https://mcp.vincia.io. Add it to your LLM client and 31 contributor tools appear in your chat — scaffold, build, preview, simulate, lint, AND publish your contribution to the marketplace, all without leaving the conversation. Sign in once via the connector's OAuth flow and your contributor identity is cached client-side.
One-click add
Pick your client. The buttons below open your client's connector flow with the URL pre-filled. If your browser doesn't handle the deep link, use the copy-button below and paste manually.
What you get
The public surface exposes 38 tools from Vincia's contributor MCP and 8 docs resources covering the contributor platform end-to-end. As of G-S93 the loop is fully chat-first end-to-end: scaffold → build → preview → simulate → publish → see install counts, earnings, review queue, and what to build next, all without leaving the conversation. The CLI remains available for power users who prefer it.
Catalog & lookup
vincia_list_capabilities— the five SDK capabilitiesvincia_get_interface— TypeScript types for a capabilityvincia_get_scope_catalog— all permission scopesvincia_get_resource_caps— per-tier resource ceilingsvincia_get_contract_version— active platform contract versionvincia_find_examples— published example contributionsvincia_get_plugin_source— source of a published pluginvincia_search_marketplace— query the marketplace
Validation
vincia_lint_manifest— validate against the v1.0 schemavincia_check_scopes— detect required scopes from code
Webhook testing (chat-first)
vincia_create_webhook_inbox— get a durable URL Stripe / GitHub / OAuth callbacks can hitvincia_list_webhook_events— poll an inbox for arrivalsvincia_webhook_event_to_fixture— convert a captured event into a simulate fixture
Analytics & discovery (new in G-S93)
vincia_get_my_contributions— list every published contribution with installs + MRR + status (one row per contribution).vincia_get_contribution_stats— detailed window stats for ONE contribution (installs / earnings / trajectory / error rate).vincia_get_my_earnings— total earnings + optional breakdown by contribution / tier / capability, reading the same source-of-truth query as the web UI.vincia_get_review_status— queue position + ETA for a pending submission (ETA labelled as a rolling 7-day-median estimate, not a guarantee).vincia_get_my_appeals— yank requests + suspended versions for YOUR contributions (read-only).vincia_list_underserved_capabilities(public, no auth) — capabilities ranked by demand/supply ratio. When called with a Bearer, the caller's already-published capabilities are de-ranked.vincia_get_marketplace_trends(public, no auth) — trending categories, top installed, hot solution-pack compositions.
Mutate & ship (G-S91)
vincia_publish— bundle + manifest inline; pushes to the marketplace via the same endpoint the CLI uses. Requires Bearervct_*(sign in via the connector OAuth flow). Scope expansion vs the last-approved version forces re-review automatically.vincia_bump_version— convenience wrapper. Resolves the current version, computes next (patch/minor/major, preserves pre-release suffixes), and publishes. Major bumps requirebreaking_change=true.vincia_edit_scopes— apply an add/remove diff tomanifest.scopesand ship a patch-bump. Added scopes trigger re-review; removed-only edits ship live.vincia_edit_metadata— update marketplace display fields (description, readme, tags, icon, doc/support URLs) WITHOUT bumping the version. Typo fixes don't pollute the version history.
Scaffold, build & preview (G-S90)
vincia_scaffold_contribution— generate a new connector / widget / blueprint / pack workspace inline (files returned as base64 in the response)vincia_render_widget_preview— accept a widget manifest + bundle bytes inline, host atsandbox.vincia.io/widget/<id>/vincia_render_blueprint_preview— host a T0 blueprint preview atsandbox.vincia.io/blueprint/<id>/vincia_render_pack_preview— host a T4 solution-pack preview atsandbox.vincia.io/solutionpack/<id>/vincia_render_composition_preview— cross-tier composition (pack + connectors + simulate chain)vincia_render_simulation— run a workerd simulation and persist the trace (server-side workerd not yet wired; returnsstatus:"unavailable"on the remote path until G-S91)vincia_simulate_run— single connector/method invocation; same workerd limitation on the remote pathvincia_list_preview_drafts— Bearer-authenticated listing of your active sandbox drafts; anonymous calls returnNOT_AUTHENTICATED
Preview-call boundary: bundle bytes are accepted as base64 in the MCP tool arg, capped at 5 MiB raw. Oversize calls return BUNDLE_TOO_LARGE at the wrapper, not at the backend. Outbound calls from preview-mode connectors stay rejected with PREVIEW_MODE — same boundary as the stdio path.
Studio-owner draft tier (new in G-S95)
Studio users (app owners and editors) can branch a draft, make changes via chat against the draft instead of live, preview the draft at a sandbox URL, then atomically publish — or discard. Available on the studio-owner surface at mcp.studio.vincia.io (separate vst_ Bearer; the same vst_ OAuth scope covers both live and draft tools, no re-auth needed).
vincia_studio:create_draft_branch— fork the live app's schema into a new draft. Max 3 active drafts per app; 7-day TTL refreshed on each preview_draft call (hard cap 30 days from branch).vincia_studio:list_my_drafts— list your active drafts across apps.vincia_studio:preview_draft— provision (or refresh) the preview app atpreview-<draft_id>.sandbox.vincia.io. Reuses the G-S79 W2 sandbox-runtime fleet — same Caddy block, same teardown cascade.vincia_studio:publish_draft— atomically flip live to the draft's accumulated schema in one DB transaction. ReturnsLIVE_DIVERGEDwith a conflict diff if live moved since branch.vincia_studio:discard_draft— throw away a draft without affecting live. Idempotent.- Every G-S82 mutating tool (
add_widget,add_collection,apply_theme, etc.) now accepts an optionaldraft_idargument — when set, the mutation lands in the draft, not live.
Onboarding prompts
Six guided prompts (audit, scaffold, simulate, lint, search, publish) walk an LLM through a first contribution — useful as self-directed onboarding inside any client.
Docs resources
The MCP resources/list exposes 8 docs sections at vincia://docs/<slug>: preview map, sandbox previews, simulations, webhook inboxes, CLI reference, MCP tools, auth + rate limits, troubleshooting. Ask the LLM and it pulls the relevant section automatically.
The CLI is still available for power users
As of G-S91, every contributor action — including publish, version bumps, scope edits, and metadata edits — works from chat. The CLI remains an option if you prefer working from a terminal or have a large bundle (>5 MiB raw) that doesn't fit in an MCP tool argument:
curl -fsSL https://get.vincia.io/install | bash vincia login # one-time, saves a contributor token vincia publish ./path # mutates the marketplace
Bundles up to 5 MiB raw go inline through the chat-first path; larger bundles fall back to the CLI for now (the upload-token path that lifts this cap is queued for a future release).
If your LLM client caches the tools/list response per session, disconnect + reconnect the connector after a server upgrade to pick up new tools (the 4 mutating tools appear after upgrading to a G-S91+ server).
Try these prompts
- Browse the catalog
Use the Vincia connector. List the SDK capabilities, then find me three payment-connector examples.
- Validate a manifest
Lint this connector manifest against Vincia's v1.0 contract: {paste your manifest.json here}.
- Look up the rules
What are the per-tier resource caps in Vincia? Show me T1 specifically.
- Test a webhook handler (chat-first)
Create a webhook inbox labelled 'stripe-test'. I'll paste the URL into Stripe's dashboard. When events arrive, convert one into a vincia simulate fixture so I can replay it.
- Build + preview a widget end-to-end (new in G-S90)
Scaffold a T3 widget called 'tip-jar' that shows a 'Buy us a coffee' button. Build the bundle in chat, then render a preview at sandbox.vincia.io using vincia_render_widget_preview and give me the URL.
- Preview a blueprint (chat-first)
Here is a T0 blueprint manifest with one collection and one workflow: {paste blueprint.json}. Use vincia_render_blueprint_preview to host it at sandbox.vincia.io and give me the admin_url so I can poke at it.
- Publish a connector — full chat-first loop (new in G-S91)
Scaffold a T1 Stripe payments connector called 'stripe-checkout'. Once the code looks good, render a sandbox simulation with vincia_render_simulation to confirm it works against the mock Stripe fixture. Then call vincia_publish with the manifest and the built bundle to ship it to the marketplace. Show me the marketplace_url when done.
- Ship a bug fix without opening a terminal
I just fixed the retry logic in my 'connector-acme' contribution. Use vincia_bump_version with bump='patch' and changelog='Fix retry loop on 429 responses' to publish the fix. Confirm the new version_id and whether reviewers will need to re-check it.
- Edit marketplace metadata in chat
The description for my 'connector-acme' has a typo. Use vincia_edit_metadata to update the description to 'A Stripe-style payments connector with idempotent retry semantics.' Confirm that no new version was created (this should be a metadata-only edit).
- How are my contributions doing? (new in G-S93)
Use vincia_get_my_contributions to list everything I've published. Then for my top earner call vincia_get_contribution_stats with window=90d and tell me whether it's growing or declining. Finally call vincia_get_my_earnings with breakdown_by=tier and summarise the mix.
- What should I build next? (new in G-S93)
I want to find an under-served opportunity. Call vincia_list_underserved_capabilities (tier T1 only) and vincia_get_marketplace_trends in parallel, then suggest 2 specific contributions I should build based on the demand/supply ratio AND what would compose well with my existing portfolio (which you can fetch via vincia_get_my_contributions). Give me MRR estimates and a quick scaffold plan.
- Where is my submission in the review queue? (new in G-S93)
Call vincia_get_review_status with slug='connector-acme'. Tell me my queue position, the ETA, and whether it's in_review yet. The ETA is an estimate based on rolling 7-day median review time — surface that nuance, do not promise a precise hour.
Subscription tiers (heads-up, not a block)
- Claude.ai web — Connectors require a Pro or Team plan. Free users hit a paywall on the Connectors page.
- ChatGPT web — Custom MCP connectors require Plus or Pro AND Developer mode enabled under Settings → Apps → Advanced settings (ChatGPT marks third-party connectors as “ELEVATED RISK”; that's expected).
- Claude Desktop — Free tier supports remote MCP connectors. The smoothest free path.
- Cursor / Windsurf / VS Code Copilot Chat — All accept remote MCP URLs without a plan upgrade.
Server status
Liveness: https://mcp.vincia.io/health. Anonymous per-IP rate limits: 60 read-only tool calls per minute, 5 heavy calls per hour. Limits emit as tool-result envelopes (not HTTP 503) so your client renders them inline.