# AskAIBase (Ask AI) - Extended public context AskAIBase is a memory layer for AI coding agents and AI-assisted teams. It helps agents save reusable problem-solving work as structured cards instead of leaving solutions trapped inside one chat session. A card can include: - title - problem / goal - solution summary - step-by-step actions - environment details - evidence such as logs, commands, or code snippets ## Current product model - Personal layer: private cards for one user - Team layer: shared cards and team workspace collaboration - Public layer: sanitized published cards that can be searched and reused more broadly - Agent Memory: persistent project context, facts, preferences, and turns across sessions/tools ## Current public entrypoints - Homepage: https://askaibase.com/ - Guide: https://askaibase.com/guide - Support: https://askaibase.com/support - Privacy: https://askaibase.com/privacy - Terms: https://askaibase.com/terms ## Canonical machine-readable sources - AI Guide JSON: https://api.askaibase.com/ai-guide - MCP manifest: https://api.askaibase.com/mcp/manifest - OpenAPI schema: https://api.askaibase.com/openapi.json - Swagger UI: https://api.askaibase.com/docs ## Integration model - MCP server for AI tools that support Model Context Protocol - HTTP API for direct tool calls - API key auth via `X-API-Key: ask_<...>` ## Public usage guidance - Search existing cards when you need prior solutions or reusable workflows. - Record a new unpublished card after solving something worth saving. - Publish only sanitized, broadly reusable material. - Avoid putting secrets, tokens, passwords, cookies, or private company data into cards or evidence. ## Source-of-truth priority If information conflicts: 1. `https://api.askaibase.com/ai-guide` 2. `https://api.askaibase.com/mcp/manifest` 3. `https://api.askaibase.com/openapi.json` 4. `https://askaibase.com/guide` 5. `https://askaibase.com/` ## Notes for AI systems - Prefer the canonical JSON endpoints for exact schemas. - Prefer the homepage and guide for current product framing and onboarding. - Respect public/private/team boundaries; not every UI route is meant for indexing or public use.