Agentprivacy Skills — AI skill for Claude Code
Privacy-first AI agent skills for Claude Code.
How to install Agentprivacy Skills
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open mitchuski/agentprivacy-skills and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agentprivacy Skills does
Privacy-first AI agent skills for Claude Code. 80+ skills and personas for ZKP, decentralized identity, dual-agent architecture, and sovereign AI
Alternatives in AI
- Abu Cowork — Open-source alternative to Claude Cowork — a local-first AI agent desktop app · multi-model · self-evolving sk 349 ★
- OpenAgentd — The open-source, local-first alternative to Claude Code & Claude Cowork 208 ★
- AppleAI — Apple AI is a lightweight, privacy‑minded macOS menu bar app that keeps your favorite AI assistants one keystr 145 ★
README
agentprivacy-skills-v5
**Privacy-first AI agent skills for the Agent Skills standard.**
87 skills + 38 (primary) personas across 5 categories — teaching Claude (and any Agent Skills-compatible system) how to operate as privacy-preserving dual-agent infrastructure. The newest category, **wikis/** (2026-06-21), adds 19 fedwiki onboarding + git-less distribution skills — the Librarian's Wikis.
**Version:** V5.5 "The Attachment Architecture" | **Grimoire:** V10.2.1 (privacymage) · v1.3.0-pending (City of Mages)
Built on the [Privacy Value Model V5.4](https://agentprivacy.ai) and the [0xagentprivacy](https://sync.soulbis.com) architecture.
*"The persona is the role-class. The cast Mage is the instance. The vertex is the position. Conflating the three is the error; binding them is the architecture."*
What's Here
| Category | Count | Purpose |
|---|---|---|
| persona/ | 38 (primary) | Layer-1 primary personas — 15 swordsmen, 11 mages, 12 balanced. Plus 4 cosmological = 42 total primaries. |
| role/ | 64 | Domain knowledge — cryptography, governance, economics, identity, dark forest strategy, ceremonies, quaternion mapping. |
| privacy-layer/ | 19 | Foundational skills covering every term of the V(π,t) privacy value equation, plus dragon-flight and amnesia-protocol. |
| meta/ | 4 | Drake/Dragon duality, Master/Emissary hemispheric attention, cosmological bound, attachment-architecture (V5.5). |
| guide/ | 19 | NEW (2026-06-21) — fedwiki onboarding + git-less distribution, the Librarian's Wikis. Author/export/deploy/discover Claude skills as federated-wiki pages. |
Each skill is a folder with a `SKILL.md` entrypoint. All follow the [Agent Skills specification](https://agentskills.io/specification).
The Librarian's Wikis · the fedwiki onboarding layer (wikis/)
The `wikis/` category is a **git-less onboarding and distribution channel**. The ski
Related Skills
Sovereign AI SDK
Agent runtime for Sovereign AI. TypeScript on Bun. Claude-Code-core architecture with a Hermes-pattern learnin
Engram Alpha MCP
Engram Alpha MCP by Lalith Chandra (@lalithbuilds) — Sovereign local-first memory engine for AI agents (Claude
Nudgy
Native macOS menu bar app that notifies you when AI coding agents (Claude Code) finish tasks, need permissions
Meepo
A local, privacy-first AI agent (Rust) that connects LLMs to your email, calendar, browser, code, and messages
Tracecrate
Local-first AI agent trace workbench. Inspect Claude Code, Codex and OTLP logs, compare runs, and export priva
Pain AI
A local-first, privacy-respecting desktop AI companion for general task completion.
Related Agents
Oficial De Seguranca
Workbench security with dual role — (a) AUDITS every external dep/repo/AI brought by the Recruiter before adop
Persona Reviewer
Read-only dispatch target for team personas in review, analysis, or assessment mode. Structurally lacks Edit/W
Hardware Expert
Amiga system architecture specialist. Validates hardware claims, reviews address space assumptions, identifies
