Unsafe retry rules
Retrying permanent client errors can create load without making delivery more likely.
Webhook retry API
Use a bounded webhook retry API with stable idempotency, destination verification, receipts, and an explicit recovery path.
A generic HTTP retry can make a temporary outage worse. Retrying every response wastes capacity, while generating a new event ID on every attempt can create duplicate contacts and notifications. A reliable policy must distinguish temporary failures, preserve event identity, and record the final destination outcome.
Form, agent, or automation -> retry API -> CRM or public HTTPS webhookRetrying permanent client errors can create load without making delivery more likely.
A repeated request without stable identity may create the same contact or task more than once.
A retry loop without a receipt leaves operators unable to prove which attempt succeeded.
Implementation
Keep the tools that create and process the lead. Standardize only the fragile handoff between them.
Open sandbox instructions →https://leadproof.jessesay.chatgpt.site/api/v1/leadsAuthorization: Bearer lp_live_your_key
Idempotency-Key: stable-source-lead-id
Content-Type: application/json
{
"destination": "https://your-crm.example/webhook",
"name": "Alex Morgan",
"email": "alex@example.com",
"source": "webhook api"
}Questions
LeadProof retries network failures, 429 rate limits, and server errors. It does not repeatedly send most permanent 4xx responses.
The production endpoint uses up to three total attempts for an eligible new delivery.
Resubmit the identical body and destination with the original Idempotency-Key and X-LeadProof-Replay: true. LeadProof verifies the stored fingerprints first.
Get an explainable risk score, prioritized fixes, and the right LeadProof plan.