Let agents act on any system without pre-built tools

When you build an in-app agent today, every action it can take has to be defined in advance. Someone maps the endpoints, builds the connector, names the tool. The agent is only as capable as whatever you pre-built for it.

That’s a hard ceiling on your agent’s capabilities. superglue removes this ceiling.

What this unlocks

A user can ask “get my last 5 invoices” or “update the billing email for customer X”, and the agent assembles the right API call for each, using the same superglue system. No one had to anticipate either request.

You always know what the agent did

Every runtime execution creates a run record. You see the full tool config, which APIs were called, what came back, success or failure, and timing. Same dashboard whether the tool was pre-built or assembled on the fly.

Secrets never leave superglue

The agent references credentials with <<placeholder>> syntax, resolved server-side at execution time. API keys, OAuth tokens, database passwords: all stay in superglue. OAuth tokens refresh automatically.

Control what each agent can reach

Restricted API keys limit access to specific systems. User-owned credentials control per-user reach. Environment isolation separates dev and prod. All enforced at the system level.

How it works under the hood

Your agent uses the superglue CLI (sg) and the superglue skill. This skill teaches your agent all superglue CLI commands, auth patterns, tool schemas, and data flow conventions.

Discover available systems
Terminal window
```
sg system list
```
Look up endpoint documentation

The agent queries real API docs stored in superglue rather than guessing endpoint shapes.
Terminal window
```
sg system search-docs --system-id stripe -k "list invoices"
```

Run an inline tool config

A single command, no persistence. No build, no save, no draft files.

sg tool run --config '{
  "id": "get-invoices",
  "instruction": "Fetch recent Stripe invoices for a customer",
  "steps": [{
    "id": "list-invoices",
    "config": {
      "systemId": "stripe",
      "url": "https://api.stripe.com/v1/invoices?customer=<<customerId>>&limit=5",
      "method": "GET",
      "headers": {"Authorization": "Bearer <<stripe_api_key>>"}
    }
  }]
}' --payload '{"customerId": "cus_abc123"}'

The agent returns the result to the user and moves on.

Setup

Install the CLI and Skill

Follow the CLI + Skills guide to install the CLI and skill for your agent framework.
Configure API access

Set environment variables for your agent’s runtime:
Terminal window
```
export SUPERGLUE_API_KEY="your-api-key"
export SUPERGLUE_API_ENDPOINT="https://api.superglue.cloud"
```
Limit what the AI agent can access with superglue RBAC roles and direct sharing.
Register your systems

Set up the systems your agent should be able to reach: APIs, databases, file servers.
Terminal window
```
sg system create --name "Stripe" --template stripe \
  --credentials '{"api_key":"sk_live_xxx"}' \
  --docs-url "https://docs.stripe.com/api"
```
See Creating a system for more details.
Instruct your agent

Add detailed instruction on which systems to use, which tools to run and which CLI commands to invoke to your agent’s system prompts and Markdown reference files.

Next steps

CLI + SkillsCLI installation, skill setup, and project-specific agent configuration

CLI CommandsFull reference for `sg tool run --config` and all other commands

Access ControlControl tool and system access with roles, sharing, and viewer/editor permissions

Run HistoryMonitor and debug all tool executions, including inline runs