canonical: https://jentic.com/apis/actiondock.app/actiondock

# Actiondock App ActionDock API

The ActionDock API puts owner approval and deterministic policy between an AI agent's tool call and a real external write. An agent submits a bounded action, which waits as an awaiting-approval job until the workspace owner approves the exact request, and ActionDock then executes it with server-held credentials the agent never sees. It exposes operations to list the available actions, track jobs and their events, read reconciliation evidence for writes whose outcome is unknown, and inspect workspace connections, billing, and usage. The credentials for each downstream API stay on ActionDock's servers, so the agent proposes writes but never holds the keys that perform them.

## For AI agents

Submit bounded API writes for owner approval, track each job to delivery, and read reconciliation evidence, across 29 operations secured with a workspace API key.

## Scope

Does not execute payments, send messages, or provision infrastructure itself. Use for approving and governing an AI agent's external API writes only.

## Capabilities

- Submit a bounded action and hold it as an awaiting-approval job until the workspace owner approves the exact request
- List the actions a workspace exposes and the approved API connections available without their credentials
- Track each job's status and read owner-reported reconciliation evidence for a write whose outcome is unknown
- Follow job and callback events from action submission through delivery and retry
- Read report artifacts and the workspace billing, usage, and storage snapshot

## Use cases

### Owner-Approved Agent Writes

Put a human in the loop before an AI agent performs a consequential write. The agent submits a bounded action to ActionDock, which holds it as an awaiting-approval job until the workspace owner approves the exact request, then runs it with server-held credentials. This bounds what an autonomous agent can trigger without ever handing it the downstream API keys.

Example prompt: Submit a bounded action, then poll the returned job until the owner approves or rejects it

### Job and Reconciliation Tracking

Follow a submitted action from proposal to delivery. The agent lists recent jobs, reads a single job's status and events, and when a write's outcome is unknown it reads the owner-reported reconciliation evidence instead of guessing. Teams use this to audit exactly which agent action produced which external effect.

Example prompt: List recent jobs, fetch one job by id, and read its reconciliation evidence for an execution-unknown write

### AI Agent Integration via Jentic

An AI agent discovers and calls ActionDock through Jentic without handling the workspace key directly. The agent searches by intent, receives the matching operation and its input schema, and submits a bounded action, while Jentic injects the stored credential at execution time. This lets the agent propose approved writes without reading the reference docs or holding secrets.

Example prompt: Search Jentic for 'submit an action for approval', load the operation schema, and submit the action with Jentic-injected credentials

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | `/v1/actions` | List available actions |
| POST | `/v1/actions/{action}` | Submit one bounded action for approval |
| GET | `/v1/jobs` | List recent jobs and workflow metadata |
| GET | `/v1/jobs/{jobId}` | Get a single job |
| GET | `/v1/jobs/{jobId}/reconciliations` | Read owner-reported evidence for an unknown write |
| GET | `/v1/connections` | List approved API connections without credentials |

## Key resources

- **Actions** — List the available actions and submit one bounded action for approval
- **Jobs** — List jobs, read a single job, and read its reconciliation evidence and events
- **Connections** — List approved downstream API connections without exposing their credentials
- **Artifacts and Billing** — Read report artifacts and the workspace billing, usage, and storage snapshot

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 47 / 100
- **Maturity:** Foundational
- **Dimensions:**
  - Foundational Compliance: 82 / 100
  - Developer Experience & Jentic Compatibility: 61 / 100
  - AI-Readiness & Agent Experience: 22 / 100
  - Agent Usability: 94 / 100
  - Security: 55 / 100
  - AI Discoverability: 64 / 100
- **View full report:** https://jentic.com/apis/actiondock.app/actiondock/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring ActionDock by hand means creating a workspace API key, sending it as a Bearer token on actiondock.app, and separately handling the owner-approval callbacks and job polling yourself. Through Jentic you install once, import ActionDock from the API Directory, store the key once, and your agent calls it.
- **Permission scoping:** ActionDock already gates every external write behind owner approval, and through Jentic you choose which of its own operations your agent may call. You can allow it to submit actions and read job status while leaving artifact deletion and the owner-only dashboard operations out of the set.
- **Credential handling:** Your ActionDock workspace API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'submit an action for approval' or 'check a job status', and Jentic returns the matching ActionDock operation with its input schema so the agent calls the right endpoint without reading the reference docs.

## Related APIs

- **Okta** — Enterprise access management and policy enforcement, versus ActionDock's per-request approval of agent writes.
- **Auth0** — Authentication and token issuance for applications, versus ActionDock's approval layer for agent actions.
- **Stripe** — A high-stakes write API whose calls ActionDock can gate behind owner approval.

## FAQ

### What authentication does the ActionDock API use?

The ActionDock API authenticates agent operations with a workspace API key sent as a Bearer token in the Authorization header, per its OpenAPI spec. The owner-only dashboard endpoints use a separate signed-in session cookie rather than an API key, and a few endpoints such as the health check and agent manifest are public. Through Jentic the key is stored encrypted by your own Jentic One instance and injected at call time, so it never reaches the agent.

### Is there an ActionDock MCP server?

You do not need a separate MCP server to give your agent ActionDock. Jentic connects it directly from the API Directory: import ActionDock, store your workspace key once, and your agent calls its approval and job operations on demand. Operations are discovered as needed, so your agent's context is not filled with tool definitions it will not use.

### Can I require owner approval before my agent performs a write with the ActionDock API?

Yes. Your agent submits the write as a bounded action, ActionDock holds it as an awaiting-approval job, and the workspace owner approves the exact request before ActionDock executes it with server-held credentials. The agent can list available actions and poll job status, but it cannot complete a write the owner has not approved.

### Can I limit what my agent is allowed to do with the ActionDock API?

Yes. Because you run Jentic One yourself, your rules decide which ActionDock operations the agent may call. You can allow it to submit actions and read job status and events while withholding artifact deletion and the owner-only dashboard operations, and every call it makes is logged.

### What are the rate limits for the ActionDock API?

The OpenAPI spec does not specify rate limits for the ActionDock API. Check the ActionDock documentation at https://actiondock.app for current limits. Through Jentic, calls run from your own instance, so you can add your own pacing around them.

### How do I submit an action for approval with the ActionDock API through Jentic?

Search Jentic for 'submit an action for approval' to get the matching ActionDock operation and its input schema, then your agent submits the bounded action and polls the returned job for the owner's decision. To run it on your own infrastructure, install Jentic One from its GitHub repo.