SPIF
Signed, tamper-evident provenance for AI outputs
What it does
SPIF packages a prompt, model metadata, tool calls, retries, and agent handoffs into one signedspif artifact you verify offline, no network trust service needed. CBOR-chunked binary, Ed25519 sign/multi-sign, checksum tamper detection. Benchmarked against real provenance systems: vs in-toto, roughly tied on verify, 8% smaller payload. Vs Sigstore keyless, ~417x faster verify (no network/transparency-log round trip). Open source, Apache-2.0.
Binary provenance format for AI outputs: ed25519 signing, audit chains, confidence metadata, EU AI Act Article 50 support, exports to PROV-JSON and OpenTelemetry. - intelogroup/spif
SPIF is an open binary format for carrying signed, tamper-evident provenance across models, tools, agents, and organizational boundaries. AI responses rarely come from a single step. They may depend on a prompt, model snapshot, retrieval result, tool call, retry, human approval, or another agent. Once that context is separated from the response, it becomes difficult to answer: SPIF packages the answer into a portable .spif artifact that can be checked locally without a network verification service. cd spif-py pip install -e " .[dev] " spif --help Rust engine cd spif-rust cargo build --release cargo test Try it in under a minute The repository includes valid, tampered, and corrupted fixtures…from github.com
Does the same job
all alternatives →
- FVFormally Verified Leaderless Log Protocol for KafkaApr 2026 · github.com · ▲5
We open-sourced the TLA+ and Fizzbee verified spec behind Ursa's storage engine. Verification across ~200K states caught a design bug that years of production missed. We then handed the spec to Claude Code — it produced a working Rust implementation (concurrent producers, compaction, fencing) without back-and-forth. We think verified specs are the best harness for coding agents: open-source the spec, let anyone implement it.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, August 2026
the whole month →- TL
Life & fun · 10d ago · louisabraham.github.io
- SA
Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…
Life & fun · Aug 2026 · toneyalexander.github.io
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com