flowai-skill-setup-ai-ide-devcontainerlisted

# AI Devcontainer Setup Creates a `.devcontainer/` configuration for AI-agent-driven development. **Architecture**: VS Code or Cursor **opens** the devcontainer (they support the devcontainer spec natively). AI tools work **inside** the container in two modes: - **VS Code extensions** (e.g., `anthropic.claude-code`, `github.copilot`) — installed automatically via `customizations.vscode.extensions`, share the same container config and env vars - **CLI/TUI tools** (e.g., `claude` CLI, `opencode` CLI) — run in the container terminal, use the same `~/.claude/` or `~/.config/opencode/` config Both modes share config directories, env vars, and global skills. This skill configures all layers. ## Prerequisites - Project root is identifiable (has `package.json`, `deno.json`, `go.mod`, `Cargo.toml`, `pyproject.toml`, or similar) - User has confirmed they want a devcontainer ## Workflow ### Step 1: Detect Project Stack Scan the project root for stack indicators: | Indicator File | Stack | Base Image | |---|---|---| | `deno.json` / `deno.jsonc` | Deno | `mcr.microsoft.com/devcontainers/base:ubuntu` + Deno feature | | `package.json` / `tsconfig.json` | Node/TS | `mcr.microsoft.com/devcontainers/typescript-node` | | `pyproject.toml` / `requirements.txt` / `setup.py` | Python | `mcr.microsoft.com/devcontainers/python` | | `go.mod` | Go | `mcr.microsoft.com/devcontainers/go` | | `Cargo.toml` | Rust | `mcr.microsoft.com/devcontainers/rust` | | None / mixed | Generic | `mcr.microsoft.