Connecting to CrawlbetReading the current crawl, stored matches and upcoming schedule.
~/crawlbet❯ agents watch --soccer ▌
crawlbet — soccer intelligence, on record.
❯ football in motion. every crawl on record.
Follow real upcoming matches from discovery to storage, forecasts, market prices and a recorded result. The crawl shows its work.
— specialist roles /— stored matches /1×2
CRAWL COVERAGE
Reading configured leagues…
Upcoming fixtures, provider forecasts and bookmaker odds.
[ source aware. every result on record. ]v1.0
00 / live crawl processconnecting
READING THE SCHEDULER
Preparing the crawl_
Connecting to the current job and stored football data.
—
01
fetch
reading status
discover upcoming matches
02
store
reading status
persist provider fixtures
03
analyze
reading status
forecasts + market prices
04
publish
reading status
timestamp qualifying signals
05
track
reading status
settle the complete ledger
inspect all crawl steps ↗
Reading the backend run details…
❯Reading the latest scheduled crawl. This screen only observes the process.
PERSISTENT FOOTBALL STOREwaiting for database status
fixtures—
forecasts—
odds records—
signals—
settled picks—
NEXT AUTOMATIC CRAWLreading schedule…
LAST PROVIDER SYNCreading sync history…
01 / the swarm
reading agents
— / —
Source: API-FOOTBALL Model: provider-v1
Markets: match result / 1×2 Focus: soccer / pre-match
AGENT
STATE
FOCUS
LAST RUN
Waiting for backend
❯ specialist roles share the football pipeline[0 active]
02 / activity timelineserver events
Reading recorded crawl events…
stdout ●waiting for connection
03 / upcoming match boardreading stored matches
the next whistle_
Real fixtures from the football store. Forecast and odds readiness update as the crawl advances.
⌁
Reading the upcoming match store
The board fills with real provider fixtures as soon as discovery writes them to the database.
TRACKED PAPER PERFORMANCE—
04 / signals worth a closer look
Signals appear after a fixture has a compatible forecast, current odds and a qualifying edge.
~/crawlbet❯ swarm --list
meet the swarm_
Specialist pipeline roles, with account-backed profiles for your interests.
global agents / provider pipeline
your agent profilessaved to your account
Sign in to create and save your own agent profiles.
~/crawlbet❯ signals --open
follow the signal_
The market, the provider estimate and the recorded reasoning behind every pick.
0 signalsprovider-v1 / decimal odds
No signals published yet.
Signals are timestamped estimates. Odds may change before kickoff; published prices are retained in the ledger.
~/crawlbet❯ ledger --settled --paper
every result. on record_
Published signals settled against provider results. Wins, losses and voids included.
PAPER PROFIT0.00u
WIN RATE0.0%
RETURN ON STAKE0.0%
SETTLED SIGNALS0 picks
MATCH
PUBLISHED PICK
ODDS
RESULT
NET UNITS
PUBLISHED
No settled results yet.
Paper accounting uses a fixed 1u stake per published signal. A win returns odds − 1u, a loss −1u, and a void 0u. This ledger records analysis performance; it does not place wagers.
~/crawlbet❯ man crawlbet
the field manual_
Know the source. Inspect the estimate. Follow the full ledger.
01 / crawl
Read the provider.
The scheduled backend imports football fixtures, prediction estimates and bookmaker odds from API-FOOTBALL. Four specialist agent roles organize the pipeline; these are scheduled data jobs using provider predictions.
02 / compare
Price the probability.
A provider probability is an estimate, not a promise. Raw implied probability is 1 ÷ decimal odds. Each inspector shows the raw edge and the margin-adjusted market probability.
edge = estimate − (1 ÷ decimal odds)03 / track
Keep the whole ledger.
The price and estimate are captured when a signal is published. Final fixture results determine the paper settlement. Unavailable data stays unavailable; no simulated picks are added to the live ledger.
Where does the data come from?
API-FOOTBALL supplies the football data, odds and provider predictions. The inspector names the bookmaker, model version and observation time. Coverage depends on the configured provider plan and competition. Crawlbet has not independently calibrated provider estimates.
What does spawning an agent do?
Sign in to save an agent profile with a league and focus. These profiles store your analysis interests. The shared scheduled pipeline runs the four global specialist roles; a custom profile does not deploy an independent AI model or place a bet.
How does the watchlist work?
Your watchlist is saved in your authenticated account. A failed save is reported, and the interface only marks the signal saved after the server confirms it.
What happens before an API key is configured?
The connected dashboard shows API key required, zero active metrics and empty live signals/results. You can choose the separate demo route to explore illustrative examples. Demo activity and results never enter this ledger.
Does Crawlbet guarantee winning bets?
No. A probability estimate carries uncertainty and a positive calculated edge can lose. The paper ledger measures tracked published picks; a short result history does not establish a profitable strategy.