Retrieve AI Ball’s public football match probabilities and transparent historical record without an account, API key, or betting recommendations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install aiball-match-analysis
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install aiball-match-analysis using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is AI Ball Match Analysis?
AI Ball Match Analysis lets an agent read AI Ball’s public, read-only JSON endpoints to report home, draw, and away probabilities for recorded football fixtures, match results, and the model’s open performance record. It includes matches whether the model’s most likely outcome happened or not, with aggregate comparisons against the pre-match favourite.
Part of Openclaw Skills, this skill is designed for factual, balanced match analysis—not a live-score feed or a recommendation service. It reports all three outcome probabilities, explains timing and coverage limits, treats confidence as a score rather than a probability, and does not advise users which side to choose.
AI Ball Match Analysis Use Cases
- Check AI Ball’s recorded home, draw, and away probabilities for a football match being played today.
- Compare the model’s most likely outcome with the final result of a recorded match.
- Review AI Ball’s overall record or results by confidence band, week, or league, alongside the pre-match favourite baseline.
- Find out whether AI Ball covers a particular competition using its current leagues list.
- Explain why a fixture is not available yet: readings are generally captured about 45–120 minutes before kick-off.
- Respond to questions about a team matchup without turning probabilities into a betting recommendation.
How AI Ball Match Analysis Works
- Identify the requested fixture, date, competition, and—when known—the user’s IANA time zone.
- Make a plain HTTP GET request to the appropriate public endpoint under
https://aiball.samagent.ai/api/v1, usinglang=enfor English names andtzwhere applicable. - For a fixture, fetch the relevant daily record and match team names case-insensitively, allowing partial names. Check the response date and convert UTC
kickoffAtvalues to the user’s time zone. - Explain availability accurately: an unrecorded fixture may not have reached its capture window;
upcomingmeans no final score is recorded, not necessarily that the match has not started. List such matches as “result not recorded yet” and exclude them from tallies. - Present all three probabilities in home/draw/away order, with one decimal place. Identify the highest as the “most likely outcome,” and describe close probabilities as a toss-up rather than a decisive finding.
- For performance figures, quote the model’s record alongside the pre-match favourite baseline for the same set of matches. Respect small samples and do not calculate a percentage when the endpoint returns
null. - End answers containing AI Ball figures with the open-record link:
https://aiball.samagent.ai/en/record?src=skill. If the endpoints cannot be reached, say so and do not substitute remembered or estimated figures.
AI Ball Match Analysis Setup
No account, API key, or installation is required. The skill uses public JSON endpoints with HTTP GET requests. Use the API reference at references/api.md for field details and example responses when working with an endpoint for the first time.
# Today's record; replace the time zone with the user's IANA time zone
curl 'https://aiball.samagent.ai/api/v1/record/daily?lang=en&tz=Europe/London'
# Weekly readings and results
curl 'https://aiball.samagent.ai/api/v1/record/weekly?lang=en&tz=Europe/London'
# Overall record and confidence-band summary
curl 'https://aiball.samagent.ai/api/v1/record/summary'
# Covered competitions
curl 'https://aiball.samagent.ai/api/v1/record/leagues?lang=en'
# Past matches; use a league key from the leagues endpoint
curl 'https://aiball.samagent.ai/api/v1/record/matches?lang=en&league=<key>&result=all&page=1&pageSize=20'
Use date=YYYY-MM-DD for a particular daily record or week_start=YYYY-MM-DD for a particular week. The daily endpoint may silently return today for an invalid date, so verify the date in its response. League filters require the competition’s key, not its display name. Times are UTC; convert them before reporting kickoff times.
AI Ball Match Analysis Data Schema & Taxonomy
The skill reads data from public endpoints; it does not describe a local database or generate files.
| Endpoint | Purpose and available filters |
|---|---|
/record/daily |
A day’s fixtures, readings, and results; supports lang, tz, and optional date. |
/record/weekly |
A week’s readings and results; supports lang, tz, and optional week_start. |
/record/summary |
Overall record and breakdowns by confidence band; includes favourite-baseline comparisons and snapshot-source counts. |
/record/leagues |
Current competition coverage; supports lang. |
/record/matches |
Past matches; supports lang, league key, result filter (all, hit, or miss), page, and pageSize. |
Important fields and interpretation:
kickoffAt: kickoff timestamp in UTC, ending inZ.upcoming: indicates that no final score is recorded yet; it does not establish whether a match has started.confidence: a 0–1 score used for confidence bands, not a probability.source:autoindicates a scheduled pre-kick-off capture;backfillindicates a historical reconstruction and does not prove capture timing.capturedAt: timestamp associated with the reading; use it withsourcewhen describing when data was captured.market: the pre-match favourite baseline, which should accompany every record tally.- A
nullrate for a confidence band indicates fewer than 30 matches; report the sample count instead of calculating a rate.
Coverage includes European league and cup football, selected competitions in the Americas and Asia. The Malaysia Super League and Philippine leagues are not covered. Check /record/leagues for the live list.
AI Ball Match Analysis Advanced Features
- Provides multiple record views: daily, weekly, overall, confidence-band, league, and filtered match history.
- Supports localization with
lang=enorlang=ms; pass the user’s IANA time zone withtzwhen known. Without a language parameter, names are returned in Chinese; the default time zone isAsia/Kuala_Lumpur. - Distinguishes scheduled
autocaptures from historicalbackfillentries, preventing unsupported claims that every reading was recorded before kick-off. - Uses the pre-match favourite as the direct baseline for record comparisons and treats the one-in-three random baseline as, at most, a secondary reference.
- Handles endpoint quirks explicitly: invalid dates may fall back to today, league filters require original league keys, and unrecorded results must not be counted in performance tallies.
- Enforces neutral reporting: include all three outcome probabilities, identify close contests, avoid fabricated figures, and never recommend a side or discuss wagering amounts.
SKILL.md
Loading
Related Openclaw Skills
Signature Splice
Signature Splice preprocesses signature images, arranges them in a specified order inside PDF signature areas, and verifies the completed document.
TikTok User Insight and Product Opportunity Analysis
A read-only TikTok research skill that converts public videos, creator profiles, and comments into structured data for audience insight, trend analysis, and product discovery.
Custom Kitchen Cabinet Renovation Pitfall Guide
A practical renovation advisory skill that helps homeowners plan custom cabinets, control budgets, and avoid common design, material, safety, and installation mistakes.
Live Events Ticketing Research
Research and compare public StubHub and SeatGeek resale-market event snapshots without purchasing tickets or guaranteeing availability.
Contract Interact Checklist
Contract Interact Checklist applies a repeatable preflight risk review to smart contract interactions and fails the check when required safeguards are incomplete.
Teradata to Huawei Cloud DWS Migration
A complete, safety-focused toolchain for migrating Teradata schemas, data, SQL objects, and workloads to Huawei Cloud DWS.
Featured*
