frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Cryptographic delegation receipts to close the user-to-operator agent trust gap

https://github.com/Commonguy25/authproof-sdk
2•Commomguy•2h ago

Comments

Commomguy•2h ago
WIMSE, AIP, and draft-klrc-aiagent-auth all address service-to-agent trust: how a downstream API verifies that an agent is authorized to call it. None of them address user-to-operator trust: whether the operator's instructions to the agent actually reflect what the user authorized.

The trust gap is structural. The delegation chain is User → Operator → Agent → Services. The user tells the operator what they want. The operator writes a system prompt. But there is no cryptographic record of the user's original intent at the moment of delegation. The operator becomes an unchecked intermediary. If an operator is compromised, goes rogue, or is compelled by a court order, the agent cannot distinguish that from legitimate instruction. The user has no recourse. Auditors have no evidence chain.

AuthProof fills that gap with a *Delegation Receipt*: a signed Authorization Object anchored to a decentralized append-only log before any agent action begins. Four fields: a structured scope (explicit allowlist, deny by default), hard boundaries the operator cannot override, a time window anchored to the log timestamp not the client clock, and a hash of the operator's stated instructions at delegation time. If the operator later instructs the agent differently, the hash mismatch is provable from the log.

The trust stack has three layers:

1. *Signed capability manifest* — the scope field references the hash of a signed server manifest, not the operator's self-reported schema. Tool server behavior that diverges from the manifest is detectable.

2. *Delegation Receipt* — removes trust in the operator as a faithful intermediary.

3. *Safescript execution binding* — the `executes` scope class requires the hash of a specific Safescript program's static capability DAG. Safescript is statically analyzable before execution; there's no dynamic dispatch. If the operator supplies a different program than the one the user authorized, it doesn't run.

For out-of-scope tool calls, the agent pauses and requests a micro-receipt from the user for that specific action. No silent capability expansion.

Anticipating the obvious objections:

*"Why not just use OAuth scopes?"* OAuth scopes represent the operator's request for access, not the user's authorization of a specific agent behavior. The operator controls the token request. AuthProof commits the user's authorization before the operator's instructions reach the agent.

*"Why a custom log? Why not a blockchain?"* The protocol is log-agnostic. Any tamper-evident append-only log with fork detection works. Certificate Transparency-style logs are a natural fit. Blockchain is one option; it's not required.

*"What if the user doesn't understand the scope they're signing?"* The scope is structured, not natural language — that's a deliberate protocol constraint. UI is a separate problem. The protocol guarantees that what the user signs is what the agent is bound by; it doesn't guarantee the user will read it carefully, any more than a contract guarantees the signer read it.

*"Safescript is obscure."* It is. The `executes` layer is the most ambitious part of the protocol and the least mature. It's a design stake in the ground for what the correct solution looks like; the practical adoption path for execution binding may involve other static-analysis tools.

npm: `npm install authproof-sdk` Whitepaper: WHITEPAPER.md in the repo

Interested in serious critique on the log design and the micro-receipt UX problem.

Kafka-ML: Connecting the data stream with ML/AI frameworks

https://github.com/ertis-research/kafka-ml
1•teleforce•2m ago•0 comments

Global Fuel Shortage Pushes Governments Toward Demand Controls

https://oilprice.com/Energy/Crude-Oil/Global-Fuel-Shortage-Pushes-Governments-Toward-Demand-Contr...
2•toomuchtodo•2m ago•0 comments

Bezier drawing tool w infinite zoom

https://fatih-erikli-potato.github.io/path-draw/
2•fatih-erikli-cg•4m ago•0 comments

OpenClaw Plugin for Claude Code CLI – Persistent Worker Pool with Smart OAuth

https://github.com/AlexChen31337/openclaw-plugin-claude-code
1•alexchen_dev•6m ago•0 comments

Astronauts Dedicate Moon Crater to Carroll Wiseman, Wife of NASA Commander

https://www.nytimes.com/2026/04/06/science/moon-crater-carroll-reid-wiseman.html
1•shpat•8m ago•0 comments

Training mRNA Language Models Across 25 Species for $165

https://huggingface.co/blog/OpenMed/training-mrna-models-25-species
2•gmays•10m ago•0 comments

It's Time to Take Down Your Smart Cameras [video]

https://www.youtube.com/watch?v=UMIwNiwQewQ
1•locusm•10m ago•0 comments

New Jersey cannot regulate Kalshi's prediction market, US appeals court rules

https://finance.yahoo.com/markets/options/articles/jersey-cannot-regulate-kalshis-prediction-1331...
1•pseudolus•11m ago•0 comments

ATS Resume Builder – Build an ATS friendly resume and get more interviews

https://ats-resume.com
2•chTijani•15m ago•0 comments

Physical Engineering AI – tools for mech engineers

https://github.com/010zx00x1/Awesome-Physical-Engineering-AI
1•010zx00x1•15m ago•0 comments

I set up a static blog that can be updated just by email

https://knxnts.xyz/blog-4.6.26.html
1•knxnts•19m ago•1 comments

I used a single power station to keep my off-grid cabin running

https://www.zdnet.com/home-and-office/energy/bluetti-apex-300-power-station-review/
1•y1n0•20m ago•1 comments

Exploring Psychic Powers Live TV Special (1989) [video]

https://www.youtube.com/watch?v=djKBUDg-v-M
1•nomilk•22m ago•0 comments

AI workout plan generator for Indian personal trainers (coachiq.in)

1•allgrips•24m ago•0 comments

Brevity Constraints Reverse Performance Hierarchies in Language Models

https://arxiv.org/abs/2604.00025
1•handfuloflight•25m ago•0 comments

Succession – Agent Lineage Evolution for an Agentic World

https://danieltan.weblog.lol/2026/04/succession-ale-for-an-agentic-world
1•danieltanfh95•34m ago•0 comments

Motorola USB OTG Problems (2023)

https://goughlui.com/2025/09/20/notes-motorola-moto-g84-5g-usb-otg-problems-bad-vbus/
2•goodburb•34m ago•0 comments

Gemini CLI accessed outside permited directory, then admits fault

https://twitter.com/i/status/2041317025583374376
1•yowayb•36m ago•1 comments

How Plausible Is 'Project Hail Mary'? Astrophysicists Have Thoughts

https://www.nytimes.com/2026/04/04/movies/project-hail-mary-scientific-accuracy.html
3•bookofjoe•39m ago•1 comments

Solod – A Subset of Go That Translates to C

https://github.com/solod-dev/solod
4•TheWiggles•41m ago•0 comments

Memory poisoning in AI coding agents

https://github.com/asamassekou10/ship-safe
1•asamassekou•42m ago•0 comments

Goodbye, middle managers. Hello, 'player-coaches' and 'org leads'

https://www.businessinsider.com/meta-block-managers-player-coaches-org-leads-2026-4
2•indigodaddy•46m ago•1 comments

Report: Apple has shifted 40% of planned MacBook production capacity to Vietnam

https://twitter.com/dnystedt/status/2041310255305642278
2•ilamont•57m ago•0 comments

Feynman: Open-source AI research agent

https://www.feynman.is/
2•m_kos•1h ago•1 comments

TraceFix – A simple tool to trace issues from logs faster

https://tracefix.vercel.app/
1•skillsettler•1h ago•1 comments

I made Claude slower and it changed how I use it

https://www.xda-developers.com/i-made-claude-slower-and-it-completely-changed-how-i-use-it/
1•NicoJuicy•1h ago•0 comments

Coexilia: Master Hash Manifest (v1.0)

https://archive.org/details/coexilia-master-hash-manifest-v-1.0
1•aegissolis•1h ago•0 comments

Analysis finds geometric thinking may come from wandering

https://phys.org/news/2026-04-analysis-geometric-human-math-module.html
2•pseudolus•1h ago•0 comments

You've Been Lied to About DNA Evidence [video]

https://www.youtube.com/watch?v=9okaPzpVhmM
1•gmays•1h ago•1 comments

The Hacker News Tarpit

https://www.joanwestenberg.com/the-hacker-news-tarpit/
4•sonicrocketman•1h ago•1 comments