AGENT SKILL

coderos-command-center-architect

Design terminal/control-center views without implementing unsafe actions.

License 许可
MIT
最近更新
2026-10-07
Stars
★ 1
Forks
0

01 前言 Front Matter

name
coderos-command-center-architect
description
Design terminal/control-center views without implementing unsafe actions.

03 SKILL.md 正文

CoderOS Command Center Architect

1. Purpose

Define the visual structure and data bindings for the CoderOS Command Center. Translate abstract orchestration states (like PR queues and security findings) into strict, read-only CommandCenterViewSpec JSON payloads.

2. When to Use

Use this skill when designing a new dashboard view or modifying how information is presented to the human orchestrator in the CLI or future frontend.

3. Required Inputs

  • The orchestration state to visualize (e.g., "Show me the Security Ledger").
  • The desired layout type (e.g., table, stream, graph).

4. Required Outputs

  • A docs/templates/coderos-command-center-view-template.md containing the JSON spec for the view.

5. Procedure

  1. Identify Data Source: Determine the file or memory location of the state to visualize.
  2. Define Panels: Break the view into CommandCenterPanel objects.
  3. Define Actions: If interactive buttons are required, map them strictly to CommandCenterAction objects that generate DraftWorkOrder JSONs.
  4. Enforce Read-Only: Ensure no action directly triggers execution. Set requires_human_approval=True.
  5. Output Spec: Generate the JSON representation.

6. Safety Boundaries

  • Do not implement live TUI code (no rich or textual Python code).
  • Do not expose secret files (e.g., .env) as data sources.

7. Model/Effort Guidance

  • Model: Gemini 1.5 Flash or Pro.
  • Effort: Medium.

8. Stop Conditions

Stop if tasked with creating an action that bypasses human approval.

9. Example Invocation

/coderos
Execute .agents/skills/coderos-command-center-architect/SKILL.md to design the `Security Findings` view.

04 同仓库更多 Skills