Verify AI agent action receipts offline: canonical form, Ed25519, chain rules, transparency log. SCITT profile, MCP server, browser verifier. Apache-2.0.
-
Updated
Oct 31, 2026 - Python
Verify AI agent action receipts offline: canonical form, Ed25519, chain rules, transparency log. SCITT profile, MCP server, browser verifier. Apache-2.0.
Offline, dependency-free verifier for ActaSeal signed receipts. Check a receipt without trusting ActaSeal.
One Receipt — one signed, content-free receipt per AI transaction that a stranger can verify offline, a person can challenge, and a regulator can inspect. Spec v0.2-draft, reference implementation, test vectors, coalition plan. AiGovOps Foundation incubation.
Supply Chain Integrity Transparency and Trust ledger application using Confidential Consortium Framework (CCF)
An audit layer for agent-to-agent commerce: signed trade manifests, default-deny spend policy, COSE_Sign1 receipts, epoch checkpoints, and rail-extract reconciliation. Specified in IETF Internet-Drafts. Maintained by VERAX Teknoloji (verax-ai.com).
Pask reference implementation — signed engagement receipts for physical-site events, SCITT-aligned wire format (pask-wire), attester (pask-attest), and adapters (pask-adapter).
Self-hosted governance layer between AI assistants and your real data. Deterministic PII masking, allow/deny policy, hash-chained audit, offline-verifiable Ed25519 receipts per access when a receipt sink is configured, coverage reconciliation against the DB's own query counters to surface gateway bypasses, and conformance vectors. MCP-native. MIT.
An open standard for verifiable execution provenance: a portable, third-party-verifiable record of what an agent did, in what order, and under what governance.
Independent, from-scratch implementation of the x402 acceptance-record conditions, run against the published conformance vectors
Entropy Protocol (ep1): seal AI-generated text when it is written; verify it later with a record that reveals no text, detect it with an exact false-positive rate, and seal inside an attested confidential GPU.
Proof-of-concept demonstrating cryptographically verifiable AI generation refusals. This repository shows how unsafe or illegal generations can be proven to have never occurred, using hash chains, signatures, and auditable evidence packs.
Verifiable compliance evidence: hash-chained ledgers, signed tip, hybrid Ed25519+ML-DSA-65, KMS custody, evidence packs, C2SP checkpoints + tlog-witness (split-view evidence, ML-DSA-44 cosignatures, real Rekor checkpoints verified), SCITT RFC 9943; verifiers in Python, JS, Go, Rust, Java + differential oracle. AGPL-3.0 / commercial.
Thin CLI adapter for the ratified P10 S2b replay verifier.
Constitutional governor for AI agents — deterministic authorization, unforgeable capabilities, offline-verifiable evidence
Independent clean-room verifier for draft-fassbender-scitt-time-anchor-06
Worked examples of the observed_effect v0 evidence contract: a bounded, recomputable observed-effect record and an independent consumer.
Automatic SCITT TS registration for every Inspect GSPC eval run. RFC 9943 + COSE_Sign1 + Ed25519 + RFC 9162 Merkle. MIT.
SCITT Transparency Service MVP for CSOAI measurement cards, implementing RFC 9943 (not conformance-tested). COSE_Sign1, Ed25519, RFC 9162 Merkle, offline receipt verification. MIT.
A self-improving, verifiable memory layer for AI coding agents.
To associate your repository with the scitt topic, visit your repo's landing page and select "manage topics."