frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

TrustAgentAI – Cryptographic receipts for MCP tool calls (non-repudiation layer)

3•kirillostrovsky•1h ago
MCP solves connectivity between agents and tools. It doesn't solve accountability.

When an AI agent calls execute_wire_transfer today, there's no cryptographic proof it happened, no proof who authorized it, and no tamper-evident record of the outcome. Editable logs and unverified timestamps don't cut it when real money or infrastructure is on the line.

I've been building TrustAgentAI — an open-source accountability layer that wraps MCP tool calls in a 3-phase signed receipt protocol:

Intent Envelope → Agent A signs: "I intend to call X with these exact args" Acceptance Receipt → Agent B signs: "I validated and accepted this intent" Execution Envelope → Agent B signs: "I executed it, here is the outcome hash"

Each receipt is signed with Ed25519 over a JCS (RFC 8785) canonical hash. All three are chained in a DAG ledger — causality is cryptographically enforced, not just logged. The ledger Merkle-batches entries and anchors roots externally (L2 blockchain), making retroactive tampering detectable even by a server admin.

The result is a Dispute Pack: a self-contained bundle that proves what happened, when, and who authorized it. Designed to satisfy auditors, insurers, and legal arbitrators — not just dashboards.

What's different from existing MCP security tools: ScopeGate, ai-runtime-guard, and similar tools focus on preventing unauthorized actions at the permission level. TrustAgentAI focuses on proving what happened after the fact — non-repudiation for the cases where permissions were granted but accountability is still needed. These approaches are complementary.

Technical stack:

Ed25519 via @noble/ed25519 (audited, zero-dependency) JCS canonicalization per RFC 8785 (deterministic hashing across platforms) DAG ledger with Merkle batching (each entry's hash includes parent hashes) Anti-replay via per-(did, nonce) uniqueness with TTL + 5s clock skew tolerance Risk budget enforcement (D4) at proxy level before execution HTTP server exposing /accept, /executed, /dispute/:traceId

The proxies are sidecars — they intercept existing MCP JSON-RPC traffic without changes to agents or tools. bashnpm install @trustagentai/a2a-core typescriptimport { ProxyAGateway, ProxyBGateway } from "@trustagentai/a2a-core/proxy"; import { DAGLedger } from "@trustagentai/a2a-core/ledger";

// Wrap your existing MCP tool call const result = await proxyA.forwardToolCall(mcpCall, executeTool); // result._a2a contains Intent + Acceptance + Execution envelopes // All cryptographically signed and recorded in tamper-evident ledger GitHub: https://github.com/kirbas/trustagent-a2a-protocol npm: https://www.npmjs.com/package/@trustagentai/a2a-core Protocol spec (v0.4): https://trustagentai.net/trustagentai-a2a-protocol Would love feedback on:

The JCS + SHA-256 hash target rule (spec §4) — is there a better canonicalization approach? L2 anchoring strategy — Base vs Arbitrum, and sync vs async anchoring? Whether the Dispute Pack format is sufficient for real legal/insurance use cases

Show HN: The Lottery of Life

https://claude.ai/public/artifacts/a62c4bac-3c05-4443-9d0a-50a9bd3f9d8d
1•atulvi•2m ago•0 comments

Undone Computer Science

https://www.undonecs.org/
1•y1n0•6m ago•0 comments

I Don't Know ML. Claude Does. 0.871 F1 on Predicting Linux Game Compatibility

https://getjump.me/posts/01-protondb-compatibility-ml-x-claude/
1•GetJump•6m ago•1 comments

Title: "Announcing S3syncy – Open-Source Real-Time S3 Synchronization"

https://pypi.org/project/s3syncy/
1•mtahle•7m ago•1 comments

Using Ledger, plain text accounting and a touch AI to fill in my UK tax return

https://www.jvt.me/posts/2026/02/01/ledger/
1•yesbabyyes•9m ago•0 comments

Agents over Bubbles

https://stratechery.com/2026/agents-over-bubbles/
1•abhis3798•9m ago•0 comments

Agent Package Manager

https://github.com/microsoft/apm
1•donutshop•10m ago•0 comments

Show HN: N0x – LLM inference, agents, RAG, Python exec in browser, no back end

https://n0xth.vercel.app/
1•redhanuman•15m ago•0 comments

Switzerland Built an Alternative to BGP

https://www.theregister.com/2026/03/17/switzerland_bgp_alternative/
2•jonbaer•16m ago•0 comments

The Healing Power of Social Friction

https://www.noemamag.com/the-healing-power-of-social-friction/
1•sonicrocketman•20m ago•0 comments

Cognitive overload: When processing information becomes a problem (2022)

https://www.mayoclinichealthsystem.org/hometown-health/speaking-of-health/cognitive-overload
1•marysminefnuf•21m ago•0 comments

Z AI Character Card Wiki

https://zaicc.net/
1•theyoungsir•22m ago•0 comments

Why investors won't know what to make of AI for a while

https://www.economist.com/finance-and-economics/2026/03/12/why-investors-wont-know-what-to-make-o...
1•petethomas•24m ago•0 comments

OTel-Native by Design – Back Ends That Export to Any Observability Stack

https://signoz.io/blog/otel-native-by-design/
1•dhruv_ahuja•27m ago•0 comments

Superpowers

https://github.com/obra/superpowers
1•evo_9•33m ago•0 comments

Show HN: Paste your YAML to calculate and optimize GitHub Actions costs

1•shubhamoriginx•34m ago•0 comments

No space, no power, no support – what life is like for Indian IT workers

https://theconversation.com/no-space-no-power-no-support-what-life-is-really-like-for-indian-it-w...
4•patrakov•35m ago•0 comments

Spite Driven Development

https://notes.zachmanson.com/spite-driven-development/
2•vismit2000•37m ago•1 comments

Show HN: Libfyaml 1.0.0-alpha1, a modern YAML library for C

https://github.com/pantoniou/libfyaml
1•fypanto•39m ago•0 comments

Ask HN: I built a geometric reasoning engine – what should I test it on?

2•gmeksriunas•40m ago•0 comments

SSH has no Host header

https://blog.exe.dev/ssh-host-header
28•apitman•42m ago•17 comments

Real or Slop? – Programming Languages Papers Edition

https://slop.zackg.me/
2•matt_d•48m ago•1 comments

Commandments of AI (Mindfulness)

2•xaj•51m ago•0 comments

Knowledge workers managing AI show collapsed productivity, not just a plateau

1•dfordp11•54m ago•1 comments

New iPhone app covering retro CPUs

https://apps.apple.com/us/app/opcodes/id6760205834
1•GrantMeStrength•58m ago•1 comments

AI – Assassinating Intelligence

https://yashgarg.dev/posts/ai-slop/
2•salkahfi•58m ago•0 comments

Show HN: Hat v0.7.0 – Fast, local automatic file compression and conversion

https://github.com/bittere/hat
1•_bittere•59m ago•0 comments

Meta's Omnilingual MT for 1,600 Languages

https://ai.meta.com/research/publications/omnilingual-mt-machine-translation-for-1600-languages/?...
1•j0e1•1h ago•0 comments

Show HN: Bank Parser – Convert US Bank Statement PDFs to QuickBooks-Ready Excel

https://bank-parser.com
1•zetbaur•1h ago•0 comments

Doubling down on open-access quantum computing

https://www.ibm.com/quantum/blog/open-plan-updates
1•jonbaer•1h ago•0 comments