Fingerprint & Attribution

Every response from Stacknet includes a fingerprint — a unique identifier linking to a prompt contract that traces the full provenance of the generation. This enables transparent attribution and origination tracking across the network.

What’s in a fingerprint

A Stacknet fingerprint encodes a prompt contract:


fingerprint = hash(stackId + nodeKeyId + skillIds + injectionIds + timestamp)

Component	Description
Stack ID	The Stack that originated the request
Node Key ID	The aISP node that executed the work
Skill IDs	Any registered skills invoked during execution
Injection IDs	System prompts, tools, or context injected by the Stack
Timestamp	Execution timestamp for ordering

How attribution works

Origination credits

Stacks earn origination credits for:

Activity	Credit
Bringing aISPs to the network	Ongoing revenue share
Bringing users to the network	Ongoing revenue share
Skills users create and register	Revenue when skills are reused
Models/tensors users create	Revenue when models/tensors are used
Secondary market key resales	Transaction fee
aISP task execution	Revenue share per task
User subscriptions	Revenue share
User billing top-ups	Revenue share

User credits

Users who create content also earn attribution credits:

Content type	When credited
Skills	When another user invokes the skill
Models/Tensors	When another user runs inference with the model
Social posts (remixes)	When content is remixed by other users

Verifying a fingerprint

Fingerprints are cryptographically signed by the executing aISP’s Ed25519 key:


// Response includes fingerprint
{
  "id": "chatcmpl-abc123",
  "fingerprint": "fp_a1b2c3d4e5f6",
  "attribution": {
    "stack": "stk_xxx",
    "node_key": "nk_yyy",
    "skills": ["skill_zzz"],
    "signed_by": "ed25519:public_key_hex"
  }
}