OpenChainGraph Suite · ART-648

Record Model Input Lineage

Attests attribute-level data lineage for a model's input data, which source system, which field, what transformation was applied, feeding which model run, distinct from every existing model-risk node (art-450/451/453/488/489), none of which attest where a model's input data came from, only what the model did with declared inputs. Input attributes are SUPPLIED and asserted by the caller (zero-egress); this node never fetches or validates against a live data warehouse or source system. An attribute declared without a source_system is reported as a legitimate finding (unmapped_attribute_count), never silently dropped or treated as an error. run_ref is optional, a caller with no model-run artifact yet still gets a valid lineage receipt. ECB Guide on effective risk data aggregation and risk reporting (May 2024) §3.4(3) requires complete, up-to-date data lineages on data attribute level for the risk indicators and critical data elements within scope; this node gives that its own citable artifact. SR 26-2 (effective 2026-04-17) scopes the model-risk context to conventional quantitative models only. Feeds the model-risk-lineage-pack chain (art-562) as an optional additional citation, never a required one.

attestation_mandateZero PIIClient-side only
🔒 All inputs are processed locally in your browser. No data is transmitted. Do not enter real personal data — use synthetic or anonymised inputs only.
Scope & reliance
See the manifest below for citations and scope. No formal correctness proof exists for this node yet (compute_proof_ready: deferred); see the proof-status note below the results.
Policy Parameters (JSON)
Result

      
Proof status
This node ships compute_proof_ready: deferred. No formal correctness proof has been produced for this kernel; the result above is a deterministic recompute, verified by a property-test floor, not a proven one.
Execution Hash & §4 Artifact
SHA-256 execution hash (JCS canonical, RFC 8785):

      

Ask your agent

Copy this paragraph into Claude, OpenClaw, or any MCP-aware agent to run this exact tool, with this sample, and verify the artifact.

Run the AINumbers MCP tool `record_model_input_lineage`. Task: Attest attribute-level data lineage for a model's input data, which source system, which field, what transformation was applied, feeding which model run, distinct from every existing model-risk node (art-450/451/453/488/489), none of which attest where a model's input data came from, only what the model did with declared inputs
Call it with arguments: {"policy_parameters":{}}
Verify before trusting: call `verify_execution_hash` on mcp.ainumbers.co (https://mcp.ainumbers.co/mcp) with the parameter `claimed_hash` set to the returned `execution_hash`, passing the full artifact the run returned (the object containing `policy_parameters` + `output_payload` + `execution_hash`; equivalently `policy_parameters` + `output_payload` with `claimed_hash`), not the bare hash string.
Return the ledger link https://ledger.ainumbers.co/ so a human can re-verify without contacting us.
PII rule: All inputs are processed locally in your browser. No data is transmitted. Do not enter real personal data — use synthetic or anonymised inputs only.
Open the tool with the sample prefilled: https://ainumbers.co/chaingraph/art-648-record-model-input-lineage.html#p=v1.H4sIAAAAAAAA_wECAP3_e31Dv6ajAgAAAA