Bunny better cold start

.agents/skills/bunny-cli/SKILL.md

@@ -0,0 +1,124 @@
+---
+name: bunny-cli
+description: Manage bunny.net resources from the command line (databases, DNS, Edge Storage, Edge Scripts, static sites, sandboxes, Stream video libraries and imports, authentication, and raw API requests) and query a Bunny Database from application code with `@bunny.net/database-client`. Use when working with bunny.net (pullzones, DNS zones/records, databases, Edge Storage zones and files, Edge Scripts, Magic Containers, static-site hosting/deploys, cloud sandboxes, Stream video libraries, uploads, collections, captions, and transcription, moving a video library into Bunny Stream from Vimeo, S3, Wistia, Mux, Cloudflare Stream, JW Player, or Brightcove), invoking the `bunny` CLI, or making authenticated API calls to api.bunny.net.
+---
+
+# bunny.net CLI Skill
+
+The bunny.net CLI (`bunny`) manages bunny.net resources from the command line. Use `bunny <command> --help` for full flag details on any command.
+
+## Critical: Authentication
+
+Commands require an API key. Authenticate first with `bunny login`, which opens a browser-based auth flow and stores the key in a local profile. Alternatively, set `BUNNYNET_API_KEY` as an environment variable or pass `--api-key` directly.
+
+Config is stored in (first match wins):
+
+- `$XDG_CONFIG_HOME/bunnynet.json`
+- `~/.config/bunnynet.json`
+- `~/.bunnynet.json`
+- `/etc/bunnynet.json`
+
+**When something goes wrong, check auth first**: run a quick `bunny api GET /user` to verify your key works. If using profiles, confirm the right one is active with `--profile`.
+
+## Quick Start
+
+```bash
+# authenticate
+bunny login
+
+# install this skill for AI coding tools (AGENTS.md + .claude/skills; --global for ~/.claude/skills)
+bunny skills install
+
+# make a raw API request
+bunny api GET /pullzone
+bunny api GET /user
+
+# manage databases
+bunny db create
+bunny db list
+bunny db shell
+bunny db migrations apply                             # run pending migrations/*.sql files
+bunny db migrations apply --pattern "*/migration.sql" # opt into a nested ORM layout
+
+# manage Edge Scripts
+bunny scripts init
+bunny scripts deploy dist/index.js
+bunny scripts list
+
+# manage cloud sandboxes (isolated containers with Claude Code inside)
+bunny sandbox create my-sandbox -e ANTHROPIC_API_KEY=sk-ant-...
+bunny sandbox exec my-sandbox -- bun install
+bunny sandbox url add my-sandbox 3000
+
+# manage Edge Storage
+bunny storage zones create my-zone --region DE --pull-zone
+bunny storage files upload ./photo.png --to images/
+bunny storage zones credentials my-zone --connection s3
+
+# manage DNS
+bunny dns zones create example.com
+bunny dns zones nameservers example.com               # is the registrar delegated to bunny yet?
+bunny dns records add example.com api A 198.51.100.1
+bunny dns records preset google-workspace example.com # apply a preset record set
+bunny dns records list example.com
+
+# host a static site
+bunny sites create my-site                            # provision (served at sites-my-site-<suffix>.b-cdn.net)
+bunny sites deploy ./dist                             # deploy a directory and publish it as the live site
+bunny sites domains add example.com --wait            # custom production domain
+bunny sites deployments publish --previous --force    # instant rollback
+
+# manage Bunny Stream video
+bunny stream libraries create my-library --link --save-env --no-add-videos   # link this directory, write .env
+bunny stream videos upload ./video.mp4 --title "Launch demo"
+bunny stream videos list
+
+# import a video library into Bunny Stream
+bunny stream import list --source vimeo                        # browse folders first
+bunny stream import --library 12345 --source vimeo --dry-run   # plan first
+bunny stream import --library 12345 --source vimeo --force     # hand every video to Bunny
+bunny stream import status --library 12345                     # encoding progress
+```
+
+## Decision Tree
+
+Use this to route to the correct reference file:
+
+- **Authenticate or switch profiles** -> `references/auth.md`
+- **Database management (create, list, show, link, delete, shell, studio, migrations, regions, tokens)** -> `references/database.md`
+- **Querying a database from application code (`@bunny.net/database-client`: connect, prepare, bind, batch, types, errors)** -> `references/database-client.md`
+- **DNS (zones, delegation checks, records, presets, BIND import/export, DNSSEC, logging, Scriptable DNS scripts)** -> `references/dns.md`
+- **Edge Scripts (init, create, deploy, link, stats, deployments/rollback, env vars, custom domains)** -> `references/scripts.md`
+- **Static sites (create, deploy, rollback, custom domains, GitHub Actions)** -> `references/sites.md`
+- **Edge Storage (zones, files, connection credentials, S3/FTP/HTTP, custom domains)** -> `references/storage.md`
+- **Sandboxes (create, exec, ssh, files list/cp, public URLs, persistent env vars, Claude Code auth)** -> `references/sandbox.md`
+- **Bunny Stream (libraries, player/security settings, video upload/fetch/embed/cleanup, captions, collections, encoding, transcription, smart titles/chapters)** -> `references/stream.md`
+- **Import a video library into Bunny Stream (Vimeo, S3, Wistia, Mux, Cloudflare Stream, JW Player, Brightcove; browse, dry run, resume, status)** -> `references/stream-import.md`
+- **Make raw API requests** -> `references/api.md`
+- **CLI doesn't have a command for it** -> use `bunny api` as a fallback (see `references/api.md`)
+
+## Global Flags
+
+Available on every command:
+
+| Flag        | Short | Default   | Description                                               |
+| ----------- | ----- | --------- | --------------------------------------------------------- |
+| `--profile` | `-p`  | `default` | Configuration profile to use                              |
+| `--verbose` | `-v`  | `false`   | Enable verbose/debug output                               |
+| `--output`  | `-o`  | `text`    | Output format: `text`, `json`, `table`, `csv`, `markdown` |
+| `--api-key` |       |           | API key (takes priority over profile and env)             |
+
+## Environment Variables
+
+| Variable                 | Description                                                                   |
+| ------------------------ | ----------------------------------------------------------------------------- |
+| `BUNNYNET_API_KEY`       | API key (overrides profile)                                                   |
+| `BUNNYNET_API_URL`       | API base URL (default: `https://api.bunny.net`)                               |
+| `BUNNYNET_DASHBOARD_URL` | Dashboard URL for browser-based auth flow (default: `https://dash.bunny.net`) |
+| `NO_COLOR`               | Disable colored output                                                        |
+
+## Anti-Patterns
+
+- **Forgetting to authenticate**: Run `bunny login` first. Without it, commands fail with a missing API key error. Use `bunny api GET /user` to verify.
+- **Hardcoding API keys in scripts**: Use `BUNNYNET_API_KEY` env var or `--api-key` flag instead of embedding keys. Better yet, use `bunny login` profiles.
+- **Relying on prompts in automation**: Prompts require an interactive terminal; piped answers are not supported. In scripts, CI, and agent sessions, pass every value as a flag and add `--force` to skip confirmations. A command that would need a prompt without a terminal does not wait: it fails fast with exit code 1 and names the flag to pass, so treat that as "retry with flags", never as success.

.agents/skills/bunny-cli/references/api.md

@@ -0,0 +1,79 @@
+# API Command
+
+## `bunny api` — Make raw authenticated API requests
+
+**Use `bunny api` when a CLI command doesn't exist for what you need.** Provides direct access to the bunny.net REST API with automatic authentication.
+
+```bash
+bunny api GET /user                                    # get current user
+bunny api GET /pullzone                                # list pull zones
+bunny api GET /database/v2/databases                   # list databases
+bunny api POST /database/v2/databases --body '{"name":"test","storage_region":"DE","primary_regions":["DE"]}'
+bunny api DELETE /dnszone/12345                         # delete a resource
+```
+
+## Syntax
+
+```
+bunny api METHOD PATH [--body BODY]
+```
+
+### Positional Arguments
+
+| Argument | Required | Description                                                     |
+| -------- | -------- | --------------------------------------------------------------- |
+| `METHOD` | Yes      | HTTP method: `GET`, `POST`, `PUT`, `PATCH`, `DELETE`            |
+| `PATH`   | Yes      | API endpoint path (e.g., `/pullzone`, `/database/v2/databases`) |
+
+### Flags
+
+| Flag     | Short | Description       |
+| -------- | ----- | ----------------- |
+| `--body` | `-b`  | JSON request body |
+
+## Request Body
+
+The body can come from three sources (in priority order):
+
+1. `--body` flag: `bunny api POST /path --body '{"key":"value"}'`
+2. STDIN (if not a TTY and method is not GET): `echo '{"key":"value"}' | bunny api POST /path`
+3. No body (default for GET requests)
+
+The body must be valid JSON — the command validates it before sending.
+
+## Headers
+
+All requests automatically include:
+
+- `AccessKey: {apiKey}` — from profile or `--api-key` flag
+- `User-Agent: bunny-cli/{VERSION}`
+- `Accept: application/json`
+- `Content-Type: application/json` (when body is provided)
+
+## Output
+
+- Always outputs JSON (pretty-printed with 2-space indent)
+- On error: extracts message from response fields (`detail`, `Message`, `title`) or falls back to HTTP status text
+- In verbose mode (`-v`): logs the request method, URL, body, and response status
+
+## Examples
+
+```bash
+# List all pull zones
+bunny api GET /pullzone
+
+# Get a specific pull zone
+bunny api GET /pullzone/12345
+
+# Create a DNS zone
+bunny api POST /dnszone --body '{"Domain":"example.com"}'
+
+# Pipe body from a file
+cat payload.json | bunny api POST /database/v2/databases
+
+# Use with a specific profile
+bunny api GET /user -p staging
+
+# Verbose mode to debug requests
+bunny api GET /pullzone -v
+```

.agents/skills/bunny-cli/references/auth.md

@@ -0,0 +1,127 @@
+# Authentication
+
+## `bunny login` — Browser-based login
+
+Opens a browser to authenticate with bunny.net and stores the API key in a local profile.
+
+```bash
+bunny login                # interactive login, saves to "default" profile
+bunny login -p staging     # save to "staging" profile
+bunny login --force        # overwrite existing profile without confirmation
+bunny login --api-key KEY  # no browser: verify and save the key directly
+```
+
+### How it works
+
+1. Starts a local HTTP server on a random port
+2. Opens `https://dash.bunny.net/auth/login` in the default browser
+3. Waits up to 5 minutes for the auth callback
+4. Extracts the API key and saves it to the profile
+5. Fetches user details and prints a welcome message
+
+### Flags
+
+| Flag        | Default | Description                                           |
+| ----------- | ------- | ----------------------------------------------------- |
+| `--force`   | `false` | Overwrite existing profile without confirmation       |
+| `--api-key` | -       | Skip the browser and save this key after verifying it |
+
+### Notes
+
+- If the browser doesn't open automatically, the URL is printed to the terminal
+- Exits with an error if the profile already exists (use `--force` to overwrite)
+- Uses `BUNNYNET_DASHBOARD_URL` env var if set (default: `https://dash.bunny.net`)
+
+### Headless and remote machines
+
+The browser flow needs a browser the user can see. `bunny login` detects when
+that is impossible (SSH session, CI, container, or a Unix host with no display
+server) and does not open one.
+
+- **With a terminal**: it warns why, then offers to take a pasted API key at a masked prompt, or to print the login URL plus the `ssh -L` port forward the callback needs.
+- **Without a terminal** (agents, CI, piped stdin): it exits immediately with a hint rather than hanging for 5 minutes on a callback nobody can reach.
+
+When running unattended, do not call bare `bunny login`. Either pass the key:
+
+```bash
+bunny login --api-key "$BUNNYNET_API_KEY"
+```
+
+or skip login entirely and export `BUNNYNET_API_KEY`, which takes priority over
+any stored profile and needs no config file. Keys given to `bunny login` are
+checked against `/user` first, so a bad key fails there instead of on the next
+command.
+
+With `--output json`, login prints `{"authenticated": true, "profile": "...",
+"name": "..."}` on stdout and nothing else, so the result is parseable.
+
+---
+
+## `bunny auth logout` — Remove a profile
+
+Deletes the stored API key for a profile.
+
+```bash
+bunny auth logout               # remove "default" profile (with confirmation)
+bunny auth logout -p staging    # remove "staging" profile
+bunny auth logout --force       # skip confirmation prompt
+```
+
+### Flags
+
+| Flag      | Default | Description              |
+| --------- | ------- | ------------------------ |
+| `--force` | `false` | Skip confirmation prompt |
+
+### Notes
+
+- Errors if the profile doesn't exist
+- Prompts "Are you sure?" before deleting (bypassed with `--force`)
+
+---
+
+## Custom API Endpoint
+
+The CLI defaults to `https://api.bunny.net` but supports custom endpoints for staging or internal environments.
+
+**Via environment variable** (takes priority over profile config):
+
+```bash
+BUNNYNET_API_URL=https://api.staging.bunny.net bunny api GET /user
+```
+
+**Via profile config** (set `api_url` in `~/.config/bunnynet.json`):
+
+```json
+{
+  "profiles": {
+    "staging": {
+      "api_key": "...",
+      "api_url": "https://api.staging.bunny.net"
+    }
+  }
+}
+```
+
+```bash
+bunny --profile staging api GET /user
+```
+
+**Precedence**: `BUNNYNET_API_URL` env var always wins over the profile's `api_url` field.
+
+---
+
+## Config Resolution Precedence
+
+Authentication is resolved in this order (first wins):
+
+1. `--api-key` flag
+2. `BUNNYNET_API_KEY` environment variable
+3. Profile from config file (via `--profile`, default: `default`)
+4. Built-in defaults (empty API key — commands will fail)
+
+The API base URL is resolved separately:
+
+1. `BUNNYNET_API_URL` environment variable
+2. `api_url` from the active profile
+3. Default: `https://api.bunny.net`