Terraform Cloud "No agents are available" agent pool in CI

The workspace uses Agent execution mode, so runs execute on a self-hosted HCP Terraform agent, not on HashiCorp infrastructure. The run queues but no agent in the assigned pool is connected, so it waits or fails.

What this error means

A remote run for an agent-mode workspace sits pending or fails with a message that no agents are available in the assigned agent pool.

terraform

No agents are available in the pool "self-hosted-prod" assigned to this
workspace. The run will remain pending until an agent connects.

Common causes

No agent is connected to the pool

Agent-mode runs need at least one running, registered agent in the assigned pool. If all agents are down, runs never start.

The workspace points at the wrong pool

The workspace is assigned an agent pool that has no capacity, while agents are actually running in a different pool.

How to fix it

Bring an agent online in the pool

Start a tfc-agent instance registered to the assigned pool.
Confirm it shows as connected in the agent pool page.
Re-queue the run so an available agent picks it up.

Terminal

docker run -e TFC_AGENT_TOKEN=${TFC_AGENT_TOKEN} \
  -e TFC_AGENT_NAME=ci-agent-1 hashicorp/tfc-agent:latest

Assign the workspace to a live pool

Point the workspace at the agent pool that actually has connected agents, or switch it to Remote execution if agents are not needed.

How to prevent it

Keep at least one agent connected per pool CI depends on.
Alert on agent pool capacity so a dead pool is noticed early.
Only use Agent mode where private networking requires it.