This page shows a real Outcome Certificate from --langgraph-checkpoint-trust verify against the partner quickstart contract (wf_partner, crm.upsert_contact, v3 tool_observed with langgraphCheckpoint).
The block below is a captured Outcome Certificate v1 from `contract_sql_langgraph_checkpoint_trust` verify (partner quickstart contract) so you can see checkpoint verdicts without running the CLI.
Verification report
Private verification snapshot for sharing in tickets or Slack. This URL is not indexed for search; see Security & Trust for how the site handles data.
Kind: outcome_certificate_v2
Human report
workflow_id: wf_partner
workflow_status: complete
trust: TRUSTED: Every step matched the database under the configured verification rules.
execution_path: execution_path_concerns=2; codes=MISSING_RUN_COMPLETED,NO_RETRIEVAL_EVENTS
- path_finding: code=NO_RETRIEVAL_EVENTS severity=low concern=context_quality scope=run_context codes=NO_RETRIEVAL_EVENTS
detail: No retrieval events recorded before tool observations (coarse missing-context signal).
- path_finding: code=MISSING_RUN_COMPLETED severity=medium concern=workflow_completeness scope=run_context codes=MISSING_RUN_COMPLETED
detail: No run_completed control event after tool observations (v2 graph).
run_level: (none)
event_sequence: normal
steps:
- seq=0 tool=crm.upsert_contact
declared: tool=crm.upsert_contact; intent=Upsert contact "partner_1" with fields {"name":"You","status":"active"}; parameters_digest={"fields":{"name":"You","status":"active"},"recordId":"partner_1"}
expected: table=contacts identity=[id=partner_1] required_fields=[name, status]
observed_database: rowCount=1
verification_verdict: outcome=VERIFIED; Matched the database.
observations: evaluated=1 of 1 in_capture_order
langgraph_checkpoint_verdicts:
tcp verified Production may advance under this checkpoint identity.
Machine JSON
{
"schemaVersion": 1,
"workflowId": "wf_partner",
"runKind": "contract_sql_langgraph_checkpoint_trust",
"stateRelation": "matches_expectations",
"highStakesReliance": "permitted",
"relianceRationale": "Contract verification used your registry and read-only SQL; every captured step matched declared expectations under the configured rules, and every LangGraph checkpoint rollup verdict is verified. You may treat this artifact as decision-grade for those steps, subject to your own scope and retention policy.",
"intentSummary": "TRUSTED: Every step matched the database under the configured verification rules.",
"explanation": {
"headline": "TRUSTED: Every step matched the database under the configured verification rules.",
"details": []
},
"steps": [
{
"seq": 0,
"toolId": "crm.upsert_contact",
"declaredAction": "Upsert contact \"partner_1\" with fields {\"name\":\"You\",\"status\":\"active\"}",
"expectedOutcome": "table=contacts identity=[id=partner_1] required_fields=[name, status]",
"observedOutcome": "rowCount=1"
}
],
"humanReport": "workflow_id: wf_partner\nworkflow_status: complete\ntrust: TRUSTED: Every step matched the database under the configured verification rules.\nexecution_path: execution_path_concerns=2; codes=MISSING_RUN_COMPLETED,NO_RETRIEVAL_EVENTS\n - path_finding: code=NO_RETRIEVAL_EVENTS severity=low concern=context_quality scope=run_context codes=NO_RETRIEVAL_EVENTS\n detail: No retrieval events recorded before tool observations (coarse missing-context signal).\n - path_finding: code=MISSING_RUN_COMPLETED severity=medium concern=workflow_completeness scope=run_context codes=MISSING_RUN_COMPLETED\n detail: No run_completed control event after tool observations (v2 graph).\nrun_level: (none)\nevent_sequence: normal\nsteps:\n - seq=0 tool=crm.upsert_contact\n declared: tool=crm.upsert_contact; intent=Upsert contact \"partner_1\" with fields {\"name\":\"You\",\"status\":\"active\"}; parameters_digest={\"fields\":{\"name\":\"You\",\"status\":\"active\"},\"recordId\":\"partner_1\"}\n expected: table=contacts identity=[id=partner_1] required_fields=[name, status]\n observed_database: rowCount=1\n verification_verdict: outcome=VERIFIED; Matched the database.\n observations: evaluated=1 of 1 in_capture_order\n\nlanggraph_checkpoint_verdicts:\nt\u001f\u001fcp\tverified\tProduction may advance under this checkpoint identity.",
"checkpointVerdicts": [
{
"checkpointKey": "t\u001f\u001fcp",
"verdict": "verified",
"seqs": [
0
],
"productionMeaning": "Production may advance under this checkpoint identity."
}
]
}