Update docs/design/AgentRun-CLI-Design.md

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

Author: LaynePeng

docs/design/AgentRun-CLI-Design.md

@@ -21,7 +21,7 @@ A developer wants to create a new agent from scratch and publish it to AgentCube
 After completing development, the CLI enables the following steps:
 1. `agentrun pack -f ./` Packages the agent source code and runtime metadata into a structured workspace directory, preparing it for image creation
 2. `agentrun build -f ./` Builds a container image from the workspace, compatible with AgentCube’s Kubernetes-based runtime environment
-3. `agentrun publish -f /` Publishes the built agent image to AgentCube, making it available for invocation, sharing, and collaboration
+3. `agentrun publish -f ./` Publishes the built agent image to AgentCube, making it available for invocation, sharing, and collaboration
 
 Use Case 2: Check Published Agent Status
 After publishing an agent to AgentCube, a developer may want to verify that the agent is fully registered and ready for use. The CLI provides a simple status check command: