EdgeLab
EdgeLab computes its own probability for every tennis match - ATP, WTA and Challenger - then checks it against the odds to spot the gaps. Public track record.
https://edge-lab.io/enConnect straight to this server’s public endpoint.
https://edgelab.fr/api/mcpWe add this server to your workspace, walk through sign-in once, then open Studio with tools ready to run.
Last probed Sep 14, 2026 · edgelab.fr
11tools discovered
/api/v1/capabilities
this document: routes served, parameters, fields with their tier and type, and which switches are on Returns: this document itself. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/pricing
the price grid: three tiers, two billing periods, two languages, already formatted - plus whether the founder offer is open Returns: no analysis data: public amounts, derived from the site's price grid. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/matches
a day's card: matches, tournament, odds, how they moved since opening, and WHICH matches carry a value signal. `hasValuePick` is free - it is what the badge on the public site shows; the pick itself (`valueSide`, `valueLevel`, `valueEdge`) is paid. One call answers « which are today's picks », where it used to take one /analysis per match Returns an ARRAY; the field counts below describe one item, with 37 free fields and 3 that need a paid viewer. Costs one call from your key's quota, refusals
/api/v1/matches/{id}
a match header (the public part of the site) plus the full odds trajectory Returns an object, with 8 fields, all free. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/matches/{id}/analysis
the full analysis in one call: dashboard, model, value pick, score matrix, aces and double faults Returns an object, with 26 free fields and 11 that need a paid viewer. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/players
Elo ranking of active players scoped to ONE tour - or, with `q`, a SEARCH by name across the whole player table (23 379 rows), which is the only way to identify someone outside the top of a ranking or today's schedule. Same row shape either way Returns an ARRAY; the field counts below describe one item, with 7 fields, all free. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/players/{key}
a player profile: Elo by surface, records, form, workload, serve, recent matches, record by opposing style Returns an object, with 28 fields, all free. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/h2h
two players projected onto a surface: the same dashboard as a match card Returns an object, with 23 free fields and 8 that need a paid viewer. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
/api/v1/track-record
the full value-signal log: history, P&L curves, and TWO summaries. `summary` measures the REAL scope - the published picks and nothing else; it is what the CLV, the official scope and the era breakdowns commit to. `summaryAffiche` is what the public /trackrecord page SHOWS: the whole of 2026, simulated from 1 January to 29 June then real, which is a deliberate product decision. Quote `summaryAffiche.pnl` when a reader asks what the page says, `summary.totalPnl` when they ask what the model has a
Todays Value
The value signals of the day, ready to read: today's card, then the value pick of each match that carries one. This is the parcours most clients start from, and it hides the opaque ids: you get player names, the model's probability and the edge. Costs up to 6 calls from your key's quota (one for the card, then one per match analysed). Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
Analyse Match By Names
The full analysis of a match, found by the two players' NAMES rather than by an opaque id. It looks the match up in the day's card, then calls the analysis. Use it when a human named the match ('Alcaraz Sinner tonight') and you do not have an id. Costs 2 calls from your key's quota. Costs one call from your key's quota, refusals included: a failed attempt is counted before it is judged.
Get your MCP into directories
A working endpoint is step one. Directory coverage is the coordinated launch across ChatGPT, Claude, Cursor, the MCP Registry, and community indexes.
Directory coverage for brandsEdgeLab computes its own probability for every tennis match - ATP, WTA and Challenger - then checks it against the odds to spot the gaps. Public track record.
Use the MCP endpoint listed on this page in your MCP client configuration. One-click install pills support Claude, Cursor, VS Code, and other hosts. Copy the remote MCP URL if your client needs a manual entry.
Operate EdgeLab? Verify ownership to take over this directory entry.
This server appears in the MCPBundles directory. Verify you operate it to take over the listing — name, description, logo, contact email, and skill content. We email a 6-digit code to a maintainer address your server publishes in /.well-known/security.txt or /.well-known/mcpbundles.json. Free, takes about a minute.
MCPBundles probed 11 tools on the live server. The tool list on this page reflects what was discovered at the last refresh — connect your client to see the full set available to your session.
EdgeLab may require signing in to the provider before tools can run. Connect through MCPBundles or your MCP client and complete any provider login when prompted.
MCPBundles is an independent platform built on the open Model Context Protocol standard. Not affiliated with Anthropic PBC or Claude.