Feature Request: Display Active Agents’ Status in CLI

[diegonix]

Feature Request: Display Active Agents’ Status in CLI

Description When Code Puppy is executing a task, the main agent can invoke several sub-agents to handle different parts of the workload. Currently, when these agents are running, there’s no visible feedback in the CLI — making it difficult for the user to understand what’s happening in real time.

Problem The lack of feedback leads to uncertainty about system activity. Users don’t know:

• Which agents are currently active
• Which models each agent is using
• How many tokens each agent has consumed or is currently using

This makes debugging and monitoring complex tasks more difficult.

Proposed Solution Add real-time feedback in the CLI showing all active agents besides the main one. Each entry should display at least:

• Agent name
• LLM model in use
• Total tokens (used + allocated)

Example (mockup):

[planning-agent]  GLM-4.6   tokens: 1200 / 200000 (10%)
[code-analyzer]  GPT-4o   tokens: 800 / 272000 (10%)
[test-runner]  Claude-3   tokens: 230 / 200000 (10%)

Implementation Suggestions

• Add an internal event emitter or logging hook in the agent orchestration layer to track agent invocation and token usage.
• Update the CLI renderer to display agent activity dynamically (e.g., refresh every few seconds or on status change).
• Consider an optional flag (e.g., --verbose or --show-agents) to toggle this feature for users who prefer quieter output.

Expected Benefits

• Better transparency into concurrent agent execution.
• Easier performance tuning and debugging.
• Enhanced user experience during long or complex operations.