These are open standards and tooling born from [PAICE](https://paice.work/) and released independently because interoperability is the point, not lock-in.

**Protocols & specifications**

| Project | What it does |
|---|---|
| [graceful-boundaries](https://github.com/snapsynapse/graceful-boundaries) | Spec for how services communicate operational limits to humans and agents |
| [turnfile](https://github.com/snapsynapse/turnfile) | File-based protocol for LLM agents to negotiate and collaborate without an orchestrator |
| [hardguard25](https://github.com/snapsynapse/hardguard25) | 25-character alphabet for human-friendly unique IDs |

**Agent skills**

| Project | What it does |
|---|---|
| [skill-provenance](https://github.com/snapsynapse/skill-provenance) | Metaskill for version tracking agent skill bundles across sessions, surfaces, and platforms |
| [skill-a11y-audit](https://github.com/snapsynapse/skill-a11y-audit) | WCAG 2.1 AA accessibility audit skill for Claude Code, Codex, and other AI agents |

**Knowledge as Code**

| Project | What it does |
|---|---|
| [knowledge-as-code-template](https://github.com/snapsynapse/knowledge-as-code-template) | Template for building structured, version-controlled knowledge bases that generate static sites, JSON APIs, and multi-output from markdown |
| [ai-capability-reference](https://github.com/snapsynapse/ai-capability-reference) | Evidence-backed reference for AI capabilities, pricing, and availability (for humans and agents) |
| [ai-regulation-reference](https://github.com/snapsynapse/ai-regulation-reference) | Structured reference tracking AI regulation obligations across jurisdictions and authorities |
| [virtual-meeting-reference](https://github.com/snapsynapse/virtual-meeting-reference) | Structured reference tracking platform capabilities across 22 virtual meeting tools |

**Tools & contributions**

| Project | What it does |
|---|---|
| [siteline](https://siteline.snapsynapse.com/) | Public scanner that evaluates whether a website presents a usable path for AI agents acting on behalf of humans |
| [substack2md](https://github.com/snapsynapse/substack2md) | Convert Substack content to markdown download for local library usage and consistent RAG storage |
| [paice-near-integration](https://github.com/snapsynapse/paice-near-integration) | Privacy-preserving AI collaboration scoring with verifiable inference and on-chain attestation via NEAR protocol |
| [OB1/recipe/obsidian-migration](https://github.com/NateBJones-Projects/OB1) | Contributor to Nate B Jones' Open Brain agentic memory / second brain project |