Contexts

Each context is a standalone prompt for building better AI agents. Browse as HTML or fetch the raw markdown directly.

• darkmode — UI dark mode with energy savings, eye comfort, and accessibility reasoning
  .md
• identity — AI transparency and foreveragent disclosure patterns
  .md
• accessibility — Inclusive design: screen readers, keyboard nav, ARIA, WCAG
  .md
• error-handling — Empathetic error messages, graceful degradation, recovery
  .md
• privacy — Privacy-first data handling, minimal collection, consent
  .md
• onboarding — First-time UX, progressive disclosure, welcome patterns
  .md
• conversation — Turn-taking, clarification, context retention, multi-turn
  .md
• safety — Content moderation, harmful content detection, safe defaults
  .md
• performance — Resource efficiency, lazy loading, minimal footprint
  .md
• testing — Agent behavior testing, conversation testing, edge cases
  .md
• deployment — Sovereign deployment, WebLLM, edge-containers, encrypted portability
  .md
• mcp — MCP for sovereign agents, guarded Bash tool execution, edge runtimes
  .md
• encryption — Client-side encryption, URL fragments, and zero-knowledge sharing patterns
  .md
• supply-chain — Dependency minimalism, vendoring, and build-process attack reduction
  .md
• zero-infra — Static-file architecture, optional services, and durability by default
  .md
• zero-dependencies — Fewer imports, smaller attack surface, auditable codebase
  .md
• local-inference — WebLLM, transformers.js, local RAG, and offline inference patterns
  .md
• portability — Transfer agents via links, QR, files, and removable media
  .md
• sustainability — Edge-first compute and energy-aware architecture tradeoffs
  .md
• threat-model — Trust boundaries, client-side attack surfaces, and mitigation patterns
  .md
• forever-agents — What Forever Agents are and how to use these contexts
  .md
• philosophy — Why architecture beats policy — the philosophical foundations
  .md

← Home