Time-based rules
Rules that use time functions now evaluate correctly. You can write conditions such ashour_of_day(created_at) >= 22 or day_of_week(created_at) in ("Saturday", "Sunday") and Watch applies them using the transaction’s created_at timestamp.Time functions also support in and not_in for lists of hours, weekdays, or day names.Setting conditions → · Time-based examples →Previous transaction rules
Theprevious_transaction() function now works as documented. Watch looks back from the current transaction’s created_at and checks whether a matching transaction exists in that window.You can match on table fields such as source and status, or metadata paths such as meta_data.status.Setting conditions → · Compound examples →Install Blnk Watch with one command
v0.1.2 is the first release with prebuilt binaries. Install on Linux or macOS (Apple Silicon) with:Welcome to our first release
Blnk Watch is live. You can monitor ledger transactions with custom risk rules, get a verdict for each transaction, and enforce limits, fraud checks, and compliance controls without running heavy infrastructure.Get started with Blnk Watch →WatchScript rules
Write rules in WatchScript (.ws), a domain-specific language for transaction monitoring. Define conditions on amount, source, destination, metadata, aggregates, and time windows, then set a verdict when a rule matches.Setting conditions → · Defining verdicts → · Rule structure →Transaction evaluation
Send transactions to Watch for real-time evaluation. Watch returns a consolidated risk assessment with a verdict such asallow, review, or block, plus scores and reasons from each triggered rule.How Watch works → · Inject transaction → · Get verdict →