integrity = UNVERIFIED — scenario placeholder · run agents to verify
This is outcome-based evidence — not an import-presence badge.
Every row in this chain is a real govern() decision (ALLOW / STEP-UP / DENY) emitted by the AGT policy gate and committed to an append-only, hash-linked log. Each entry's curr hash becomes the next entry's prev, so any tampering or deletion breaks the chain and fails hash chain verification. The assurance KPMG presents is that these allow/deny/step-up records happened and are intact — we are not claiming a feed exists, we are proving what it decided.
Hash chain · tamper-evident log
SHA-256 hash chain · append-onlyDENY
shadow-llm-agent · delete
prev 0x7c10a9…b3e4 → curr 0x9f3a17…42b1
12:02:58
ALLOW
grid-ops-agent · read:ot-telemetry
prev 0x4b8205…11aa → curr 0x7c10a9…b3e4
12:03:22
ALLOW
sandbox-coder · exec:sandbox
prev 0x2d55e8…7c0c → curr 0x4b8205…11aa
12:03:40
STEP-UP
finance-close-agent · write:journal
prev 0x18ef03…9a73 → curr 0x2d55e8…7c0c
12:03:55
STEP-UP
dsar-export-agent · export:pii
prev 0xc6b1f4…0e18 → curr 0x18ef03…9a73
12:04:18
ALLOW
recon-agent · read:identity-graph
prev 0x8a40d2…55cd → curr 0xc6b1f4…0e18
12:04:21
DENY
marketing-scraper-agent · read:customer-pii
prev 0x3f97ba…ee61 → curr 0x8a40d2…55cd
12:04:33
ALLOW
vault-rotator-agent · rotate:secret
prev 0xa3e9c4…d2f7 → curr 0x3f97ba…ee61
12:04:47
integrity = UNVERIFIED (scenario placeholder)
1,204 events · root 0xa3e9c4f1…d72f7 · last verified —
Chain integrity
last 24h
Total entries
1,204
Chain length
1,204 blocks
Algorithm
SHA-256 hash chain
Tamper checks passed
— / — (pending run)
Broken links
0
Root hash
0xa3e9c4f1…d72f7
—
Verification pending (no live run)
Every block re-hashed & chain-linked
source_system=AGT audit log (hash-chained)
tenant=kpmgplayground
source_mode=scenario (live after a run)
algorithm=SHA-256 hash chain
agt=4.1.0
2026-06-14