
FutPicks
io.github.KhavelPublished by GitHub account Khavelv1.0.2Updated Oct 24, 2026
Probable football lineups with start probability and expected fantasy points, graded in public.
Overview
FutPicks gives an assistant probable football lineups with start probabilities and expected fantasy points, plus matches, model picks and a public track record.
- What you can do
- Look up clubs, today's and upcoming fixtures, and match details with the projected XI and model read
- Browse today's picks, a filterable picks board and individual pick details
- Check the public track record of graded picks, filtered by market, league, rating, outcome or date
- With a data key: picks history, match evaluation, backtest runs, CSV export and a data catalog
- With an ops key: trigger and poll admin operations such as ingestion, scoring, settlement and calibration runs
- When to use it
- Useful when you want fantasy or betting-oriented football analysis for LaLiga, Premier League, Serie A, Bundesliga, Ligue 1 and Primeira Liga, especially projected starting lineups and model picks with measured accuracy. The public tools need no account, so it is easy to try before adding keys.
- Requirements
- Runs locally over stdio via npx (Node.js). Every tool calls the public futpicks.com REST API over HTTPS, so network access is needed. Public lineup, match and track-record tools need no key. Pro data tools need FUTPICKS_API_KEY (data-scope key, minted in the web app under Account, API Keys); admin ops tools need FUTPICKS_OPS_TOKEN. FUTPICKS_API_URL optionally overrides the base URL.
Installation
In SourceWeft
- Open FutPicks in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
FutPicks MCP Server
Model Context Protocol server for football data: probable lineups with a per-player start probability and expected fantasy points, match analysis, model picks and a public track record — LaLiga, Premier League, Serie A, Bundesliga, Ligue 1 and Primeira Liga.
Every tool calls the public futpicks.com REST API over HTTPS. No database access, no shell, no local secrets beyond the API tokens you put in the environment. The lineup, match and track-record tools need no key at all.
The lineups are graded in public against the confirmed XI, so the accuracy the tools report is measured, not claimed. Same data as futpicks.com/onces/la-liga.
Quick start
Add it to any MCP client (Claude Code shown here) — no install step, npx fetches it:
Then ask for a probable lineup: flab_teams to find a club, flab_matches_upcoming for the
fixtures, flab_match_details for the projected XI and the model's read of the match.
To unlock the Pro and admin tools, add the keys:
Building from source
For local development with hot reload: npm run dev.
Environment variables
Public tools work with no token at all.
Where to get the tokens
Mint API keys from the web app: futpicks.com → Account → API Keys (Cuenta → Claves de API).
FUTPICKS_API_KEY— create a Data-scope key. It looks likeflab_live_…and satisfies the Pro-tier (ProTier) endpoints. Pro and admin users can mint these.FUTPICKS_OPS_TOKEN— create an Ops-scope key. It looks likeflab_ops_…and satisfies the admin (Ops) endpoints. Only admins can mint these.
The raw key is shown once at creation, so copy it into your environment immediately. Keys are
stored only as SHA-256 hashes, support optional expiry, and can be revoked from the same screen;
revoked or expired keys are rejected with 401.
Tools
Public (no auth)
Data tools (require FUTPICKS_API_KEY)
Ops tools (require FUTPICKS_OPS_TOKEN, admin-only)
flab_ops_run kinds: ingestion-runs, scoring-runs, settlement-runs, calibration-runs,
intelligence-refreshes. The server auto-generates an Idempotency-Key per ops POST.
Claude Desktop configuration
Add to claude_desktop_config.json:
Verify locally
Then call flab_health (no token needed) — it should return live system status.
Security
- No tool reads DB connection strings, appsettings secrets, or shells into the VPS.
- All inputs are validated with zod; page sizes and ranges are capped server-side.
- Tokens are read from environment variables only and never logged.
- Data/ops tools fail with a clear message when their token is missing — they never fall back to an unauthenticated call.
- Ops mutations send a fresh
Idempotency-Keyso retries are deduplicated by the API.
Source: README.md at commit 9f57999
Tools
0Version history
1- v1.0.2LatestSep 16, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts