cli

Bilingual

Show HN: Supapool – a Supabase per coding agent in ~400 ms
2 months ago
- Supapool provides ephemeral, isolated Supabase instances for running commands in CI or agent workflows.
- The 'npx @supapool/cli run' command wraps any command, leasing a clean instance with injected Supabase credentials.
- Each instance includes real Postgres, Auth, and S3-compatible Storage, avoiding mocks and parallel agent conflicts.
- Instances are colocated near the agent, spin up quickly, and are released upon command exit with all data wiped.
- The CLI injects standard Supabase environment variables (e.g., SUPABASE_URL, DATABASE_URL) and aliases for popular frameworks.
- Leases have a default 30-minute TTL, renewed every 5 minutes; CI jobs skip browser login by using an API key.
- The npm package also offers a 'withInstance' function for programmatic use, ensuring lease acquisition and release even on errors.
`bun init` automatically creates a Claude.md file by default
2 months ago
- Bun can initialize a project with `bun init` and accept defaults using `-y` flag.
- It creates package.json, tsconfig.json/jsconfig.json, entry point (default index.ts), README.md, and optionally CLAUDE.md or Cursor rules.
- The command runs `bun install` at the end to install `@types/bun`.
- Options include `--help` (-h), `--yes` (-y) to accept prompts, and `--minify` (-m) for type definitions only.
- Project templates are available via `--react` with presets like `tailwind` or `shadcn`.
- Output files vary by template and include project configuration and scaffolding.
Why I'm building my own CLIs for agents
2 months ago
- MCP (Model Context Protocol) has significant token consumption issues, using up to 15,000 tokens for a single tool like Playwright, which limits usable context window.
- LLMs face non-linear computational scaling with context length, making long sessions expensive and slow.
- Selectively disabling MCP tools is not supported in some platforms like Claude Code, leading to frequent context length limits.
- Using CLIs instead of MCP dramatically reduces token usage (e.g., 71 tokens for Linear CLI vs. thousands for MCP).
- Building custom CLIs is straightforward: from OpenAPI specs, API docs, or even HAR files from websites without public APIs.
- CLIs enable agents to perform complex multi-step tasks (e.g., finding bug emails, creating tickets, fixing issues, drafting replies).
- Skills standardize CLI use but have challenges with dependencies and cross-platform binaries.
- Project-specific CLI instructions in git-tracked files improve organization and agent performance.
- This CLI approach achieves MCP's vision (connecting data sources to LLMs) without its token consumption problems.
- The author predicts a resurgence of terminal UIs and custom CLIs for agent workflows.

Show HN: Supapool – a Supabase per coding agent in ~400 ms

`bun init` automatically creates a Claude.md file by default

Why I'm building my own CLIs for agents