Skip to product information

Prove OpenAI Retry Cache HITs in n8n with Metrecept JWKS

Prove OpenAI Retry Cache HITs in n8n with Metrecept JWKS

 (200+Reviews)
Regular price £52.99
Regular price £52.99 Sale price
SAVE Sold out
Instant Digital Download
Unlimited Downloads
Lifetime Access in Your Account
🔥
128+ Sold
Popular with n8n builders
23 people viewing
High interest right now
9 added today
Fast-moving digital product
Prove OpenAI Retry Cache HITs in n8n with Metrecept JWKS

Prove OpenAI Retry Cache HITs in n8n with Metrecept JWKS

Regular price £52.99
Regular price £52.99 Sale price
SAVE Sold out

Prove OpenAI Retry Cache HITs in n8n (with Metrecept JWKS verification)

This n8n workflow sends the exact same OpenAI-style chat completion request to Metrecept twice—first to demonstrate a cache MISS, then (optionally) to produce and prove a cache HIT—verifying the returned receipt signature using Metrecept’s public JWKS.

What this workflow does

  • Accepts input via manual run or webhook: a POST request containing prompt or messages (optionally model and prove_retry).
  • Normalizes and freezes the payload: converts your input into an OpenAI Chat Completions request body, freezes it as a JSON string, and returns HTTP 400 if no messages are provided.
  • Sends the request to Metrecept: posts the frozen payload to /v1/chat/completions and captures cache/billing headers. Returns HTTP 402 on budget errors and HTTP 502 for other upstream failures.
  • Generates a retry HIT (optional): when prove_retry is enabled and the first response is a cache MISS, the workflow sends the identical payload again to trigger a cache HIT.
  • Verifies the receipt signature (JWKS): if a receipt JWS header is present, it fetches Metrecept’s public key directory (JWKS) and verifies the Ed25519 signature. If missing, it reports verification as unavailable.
  • Returns a consolidated proof object: includes mode, cache status, billed amounts, receipt verification result, and current usage from Metrecept /v1/usage.

Use cases

  • Audit that “retry” behavior in your OpenAI-style integrations actually yields cache HITs.
  • Validate Metrecept receipt integrity for compliance or internal reliability testing.
  • Estimate billed impact by comparing MISS vs HIT scenarios with returned usage data.

Technical details

  • n8n nodes: webhook, if, set, code, switch, plus sticky notes for clarity.
  • Metrecept integration: HTTP calls to /v1/chat/completions and /v1/usage, plus public JWKS fetch for receipt verification.
  • Auth setup: an n8n HTTP Header Auth credential using Authorization: Bearer <your_intermediate_key>.
View full details