Appearance

Connect your agent ​

smplkit is agent-native: a single Model Context Protocol (MCP) server exposes the whole platform — feature flags, configuration, logging, audit, and scheduled jobs — to any MCP-capable coding agent. Connect it once, and your agent provisions and operates every primitive in plain language. You stay the owner: every change runs under your account and shows up in your console alongside the changes you make by hand.

The server is hosted at https://mcp.smplkit.com/api/mcp and is stateless — it holds no credentials of its own.

1. Add the server and sign in ​

Point your MCP client at the server URL. The first time it connects, the client opens your browser for a one-time sign-in — Continue with Google or Microsoft (standard OAuth). After that it reconnects and refreshes access on its own; you're not prompted again unless you disconnect. There's no key to mint, copy, or rotate.

bash
claude mcp add --transport http smplkit https://mcp.smplkit.com/api/mcp

…or add it to .mcp.json:

json
{
  "mcpServers": {
    "smplkit": {
      "type": "http",
      "url": "https://mcp.smplkit.com/api/mcp"
    }
  }
}

On first use Claude Code opens your browser to sign in, then remembers the connection.

Prefer a static key? ​

For non-interactive use — CI, scripts, headless clients, or writing code with the smplkit SDKs — skip the browser and authenticate with an API key as a bearer token instead.

  1. Sign up at smplkit.com. Continue with Google or Microsoft — SSO sign-ups are email-verified instantly, so you can mint a key right away.
  2. In the console, open API Keys → Create, name the key, and copy the secret. It starts with sk_api_.

Send it on every request as Authorization: Bearer YOUR_SMPLKIT_API_KEY (a custom X-Smplkit-Api-Key: YOUR_SMPLKIT_API_KEY header is also accepted). For example, add a headers block to the config above:

json
{
  "mcpServers": {
    "smplkit": {
      "type": "http",
      "url": "https://mcp.smplkit.com/api/mcp",
      "headers": { "Authorization": "Bearer ${SMPLKIT_API_KEY}" }
    }
  }
}

The SDKs read the same key from SMPLKIT_API_KEY. A key carries the role of the user who minted it — mint the narrowest role that does the job (see API keys). Store it like a password.

2. Ask your agent ​

Once connected, express intent in plain language — the tools translate it into the right API calls:

"Create a boolean flag new-checkout, off by default, then turn it on in prod only for enterprise users."

"List my environments, then set database.host to db-staging.internal for staging."

"Raise the sqlalchemy.engine logger to DEBUG in production while I debug, then reset it."

"Schedule my cleanup endpoint to run nightly at 2am New York time, then run it once to prove it works."

The agent makes the change and proves it worked — it can fire a job on demand or dry-run a SIEM forwarder before saving — then answers follow-up questions straight from the live state.

What your agent can operate ​

One server exposes every smplkit primitive. The tools share intent-named verbs — list_*, get_*, create_*, set_*, delete_* — and hide the JSON:API envelopes, per-environment nesting, and full-replace PUTs behind partial-intent calls.

CapabilityWhat the agent can do
Feature flagsCreate flags and set per-environment values, kill switches, and targeting rules.
ConfigCreate keyed, typed config and set values per environment.
LoggingRaise or reset runtime log levels per environment.
AuditSearch the audit log and manage SIEM forwarders.
Scheduled jobsCreate, run, and monitor scheduled HTTP jobs with full response capture.
EnvironmentsList the account's environments — the valid targets for every per-environment change.

Flags, config, and logging are set per environment, so the agent calls list_environments to discover valid targets (production, staging, …) before it writes.

You stay the owner ​

The agent is increasingly the user; you remain the owner.

  • Every change runs under your account — attributed to the signed-in user (or, on the key path, to the API key) — and appears in your console next to changes made by hand.
  • Least privilege when you need it. With OAuth the agent acts with your role. To narrow what an agent can do, connect it with a scoped API key instead (see API keys) — a VIEWER key, for example, lets an agent read state but never change it.
  • Nothing is stored server-side. The MCP server is stateless and holds no credentials; whether you connect with OAuth or an API key, tokens are validated and used per request, never cached or logged.

Reaching your services ​

smplkit calls job targets and SIEM forwarder destinations from the cloud, so those URLs must be reachable from the public internet — localhost and private addresses won't fire. To target a local server during development, point at its deployed URL or expose it with a tunnel (cloudflared tunnel --url http://localhost:PORT or ngrok http PORT) and protect it with a secret auth header.