Syntax Automations · Prototype
Lead Routing Rules Auditor
Test which team receives each inquiry—and see where your rules disagree with the intended outcome.
Reviews supplied rules and synthetic examples. Does not connect to or change your CRM.
1. Load an example or edit the inputs
Fictional home-service requests demonstrate incorrect rule order, uncovered areas, and an unavailable team.
Supported format and evaluation rules
- First match wins. Lower, unique, non-negative integer priorities run first.
- All conditions in a rule must match. Every matching rule is recorded.
- Fields: service and area (strings); urgent and businessHours (booleans).
- Operators: eq for one exact value; in for a non-empty array of values.
- Comparisons are case-sensitive. Missing fields do not match. No type coercion.
- Fallback applies only when no rule matches. Unavailable teams are not automatically replaced.
- Cases without expectedDestination are simulated, not marked correct.
- Limits: 100,000 characters per editor; 100 rules; 500 cases; 100 destinations; 8 conditions per rule; 50 values per in condition; 300 characters per text value.
Demo expectations: urgent after-hours requests in north/south go to the after-hours desk; normal plumbing and HVAC requests go to their teams; electrical requests use manual review in this availability snapshot. Missing-area and unsupported-area examples expect manual review. These eight cases are illustrative, not exhaustive.
Use fictional or redacted examples only. Do not enter customer records, passwords, API keys, or confidential configuration. This tool does not connect to your CRM.
Expected destinations must express your intended policy independently of the rules. T06 intentionally omits area.
Loading the fictional example…
2. Inspect the supplied rules and cases
Previews reflect the last successful validation, not a live CRM.
3. Review the results
Run an audit to see explained results.
Example files and companion skill
The input bundle contains a config object and a testCases array. Copy those sections into the corresponding editors. The example report documents the flawed fixture.
- Download fictional example input (JSON)
- Download example findings report (Markdown)
- Download companion agent skill (Markdown)
The skill guides preparation and interpretation. The Markdown file does not execute tests by itself.
Need help turning findings into a working intake workflow?
Request a workflow review to discuss your current tools, handoffs, and implementation questions. This prototype does not inspect your live systems.
Request a workflow review