# foreveragents.dev

> A library of agent-building context prompts.

Each context is a standalone markdown file containing implementation guidance for a specific aspect of building AI agents. Fetch any context directly by its .md URL.

## Contexts

- [Dark Mode](https://foreveragents.dev/context/darkmode.md): UI dark mode with energy savings, eye comfort, and accessibility reasoning
- [Agent Identity](https://foreveragents.dev/context/identity.md): AI transparency and foreveragent disclosure patterns
- [Accessibility](https://foreveragents.dev/context/accessibility.md): Inclusive design — screen readers, keyboard nav, ARIA, WCAG
- [Error Handling](https://foreveragents.dev/context/error-handling.md): Empathetic error messages, graceful degradation, recovery
- [Privacy](https://foreveragents.dev/context/privacy.md): Privacy-first data handling, minimal collection, consent
- [Onboarding](https://foreveragents.dev/context/onboarding.md): First-time user experience, progressive disclosure, welcome patterns
- [Conversation Design](https://foreveragents.dev/context/conversation.md): Turn-taking, clarification, context retention, multi-turn patterns
- [Safety](https://foreveragents.dev/context/safety.md): Content moderation, harmful content detection, safe defaults
- [Performance](https://foreveragents.dev/context/performance.md): Resource efficiency, lazy loading, minimal footprint
- [Testing](https://foreveragents.dev/context/testing.md): Agent behavior testing, conversation testing, edge case coverage
- [Deployment](https://foreveragents.dev/context/deployment.md): Sovereign deployment — local inference servers, edge-containers, encrypted shareable agents
- [MCP Integration](https://foreveragents.dev/context/mcp.md): MCP for sovereign agents — guarded Bash tool execution and edge runtime patterns
- [Encryption](https://foreveragents.dev/context/encryption.md): Client-side encryption, URL-fragment keying, and zero-knowledge sharing
- [Supply Chain Security](https://foreveragents.dev/context/supply-chain.md): Dependency minimalism, vendoring, and deterministic integrity checks
- [Zero Infrastructure](https://foreveragents.dev/context/zero-infra.md): Static-file architecture and optional backend capability layers
- [Zero Dependencies](https://foreveragents.dev/context/zero-dependencies.md): Fewer dependencies, smaller attack surface — why minimal imports produce more resilient agents
- [Local Inference](https://foreveragents.dev/context/local-inference.md): Local-first AI — local servers as primary, in-browser as graceful downgrade
- [Portability](https://foreveragents.dev/context/portability.md): Agent transfer via links, QR, USB, and offline files
- [Sustainability](https://foreveragents.dev/context/sustainability.md): Edge-first compute and energy-aware architecture principles
- [Threat Model](https://foreveragents.dev/context/threat-model.md): Trust boundaries, extension risk, injection, and storage exposure
- [Forever Agents](https://foreveragents.dev/context/forever-agents.md): Precise definition and structured requirements for durable, privacy-first AI agents
- [Philosophy](https://foreveragents.dev/context/philosophy.md): Why architecture beats policy — the philosophical foundations