SocialClaw
Audience workflow

Use SocialClaw as an X scheduler for AI agents.

SocialClaw fits AI-agent X workflows because the workspace holds the connected X account, media handoff, provider-aware validation, and delivery state while the agent stays focused on planning.

Account type X user account Text posts, media posts, and campaign reply steps
Best fit AI agents Reply logic is provider-specific, not a universal interaction model.

What SocialClaw supports for this workflow

A quick overview of what this workflow supports.

Agent fit One workspace API key The X scheduler route uses the same workspace API key model used across AI-agent workflows.
Media rule Up to four images or one video SocialClaw uploads X media natively during publish.
Agent-safe planning Validate and publish-preview first Capabilities and publish-preview keep the agent inside current X limits before it applies a schedule.
Campaign support Reply steps are supported Explicit X reply steps are supported in campaign flows when the child step declares interaction_type reply.

What the payload needs

The main fields and constraints to account for.

  • Workspace API key auth
  • Connected X user account in the workspace
  • Text plus optional image or video assets
  • Optional replyControl and campaign reply step data

How the workflow runs

Connect the account, validate the post, then publish.

Step 1

Authenticate the agent with the workspace key

Log the agent into SocialClaw with the workspace API key instead of storing X auth directly inside the agent.

Step 2

Inspect X capabilities before planning

Read capabilities and publish-preview so the agent sees the current X media limits and any reply-control requirements.

Step 3

Validate the schedule or campaign

Validation happens before apply, especially when the agent emits media posts or explicit X reply steps.

Step 4

Apply and inspect run state

Create the scheduled run, then inspect post or run state from the same hosted workspace.

CLI and HTTP examples

Example commands and requests for this route.

CLI flow X scheduler for AI agents
socialclaw login --api-key <workspace-key>
socialclaw accounts capabilities --provider x --json
socialclaw accounts action --account-id <account-id> --action publish-preview --input x-preview.json --json
socialclaw apply -f x-agent-schedule.json --idempotency-key x_agent_1 --json
HTTP API flow Hosted SocialClaw endpoints
curl -sS \
  -H "Authorization: Bearer $SC_API_KEY" \
  "https://getsocialclaw.com/v1/keys/validate"

curl -sS \
  -X POST \
  -H "Authorization: Bearer $SC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"provider":"x"}' \
  "https://getsocialclaw.com/v1/connections/start"

curl -sS \
  -X POST \
  -H "Authorization: Bearer $SC_API_KEY" \
  -H "Content-Type: application/json" \
  -d @schedule.json \
  "https://getsocialclaw.com/v1/posts/validate"

curl -sS \
  -X POST \
  -H "Authorization: Bearer $SC_API_KEY" \
  -H "Content-Type: application/json" \
  -d @schedule.json \
  "https://getsocialclaw.com/v1/posts/apply"

Current limitations

A few practical limits to keep in mind.

  • Reply logic is provider-specific, not a universal interaction model.
  • X supports up to four images or one video per post, not both together.
  • Comment publishing is not supported.

Related routes

Related integrations and use cases.

Frequently asked questions

Why does the X route fit AI-agent workflows?

The route keeps account connection, media upload, validation, apply, and inspection in one workspace so the agent can focus on planning instead of auth and delivery plumbing.

Can AI agents use reply chains on X?

Yes. Explicit X reply steps are supported inside campaign flows.

What should an agent inspect before applying X work?

Inspect capabilities and publish-preview first so the schedule stays inside X media and interaction limits.

Run the X scheduler for AI agents workflow in SocialClaw

Connect the account once, validate before publish, and keep everything in one workspace.