Add prompt

Add a published prompt to an agent so that it is used as the agent’s instruction at runtime.

Before you begin

1. Follow the Get started guide to set up agentregistry and start the agentregistry daemon.
2. Create an agent, such as with the arctl agent init command.
3. Publish a prompt in agentregistry.

Add a prompt to an agent

1. List the prompts that are published in agentregistry.

   arctl prompt list

   Example output:

   NAME          VERSION   DESCRIPTION
   code-review   1.0.0     System prompt for code review agent
   

2. Add a prompt reference to the agent manifest (agent.yaml). The --registry-prompt-name flag specifies the prompt name that you used when you published the skill in agentregistry. The following command adds a prompts section to your agent.yaml file. You can pin a specific version by using the --registry-prompt-version flag. If omitted, the latest version is resolved at runtime.

   arctl agent add-prompt reviewer \
     --project-dir myagent \
     --registry-prompt-name code-review
   To use a prompt from a different registry, use the --registry-url flag.

   arctl agent add-prompt reviewer \
    --registry-prompt-name code-review \
    --registry-url https://registry.example.com 

   Example output:

   ✓ Added prompt 'reviewer' to agent.yaml
   

3. After adding a prompt, verify that the agent.yaml file was updated.

   cat myagent/agent.yaml

   Example prompts section in the agent.yaml file:

   prompts:
     - name: reviewer
       registryURL: http://localhost:12121
       registryPromptName: code-review
       registryPromptVersion: latest
   

4. Re-build the agent image.

   arctl agent build myagent

5. Run the agent. When you run the agent with arctl agent run, the following steps are execute:

   1. The CLI reads the prompts section from the agent.yaml file.
   2. For each prompt reference, the CLI fetches the prompt content from the registry’s REST API.
   3. The resolved prompt content is written to a prompts.json file in the agent’s config directory.
   4. The agent’s prompts_loader.py (autogenerated) reads the prompts.json file at startup.
   5. The build_instruction() function returns the resolved prompt content as the agent’s instruction, replacing the default instruction.

   arctl agent run myagent
   If no prompts are resolved (for example, the registry is unreachable or no prompts are configured), the agent falls back to the default instruction defined in the agent.py file.

6. Ask the agent what it checks during a code review. Verify that you get back the steps that you defined in your prompt.

   

   

Next

Deploy the agent to your environment.

Cleanup

To remove a prompt from the agent, edit the agent.yaml file and remove the prompt reference from the prompts list. Then, re-build and re-run or re-deploy the agent.