Feature: Add OpenClaw Agent Context Generator (Agents, Subagents, Skills, Connections)

[matheuscbs]

Feature: OpenClaw Agent Context Generator

Problem

ai-coders-context currently generates repository context focused on code understanding, but it does not support generating structured context tailored for OpenClaw multi-agent architectures.

OpenClaw requires explicit definitions for:

• Agents
• Subagents
• Skills
• Tool bindings
• Inter-agent connections

There is no generator mode optimized for this structure.

Objective

Add a new generation mode to produce structured context compatible with OpenClaw, enabling automatic scaffolding of:

• Agent definitions
• Subagent hierarchies
• Skills registry
• Tool integrations
• Agent communication flows

The output should generate a YAML or JSON structure compatible with OpenClaw.

Example:

agents:
  - name: Orchestrator
    role: Task router
    subagents:
      - ResearchAgent
      - ExecutionAgent
    skills:
      - memory.read
      - memory.write

  - name: ResearchAgent
    role: Data retrieval
    skills:
      - web.search
      - db.query

connections:
  - from: Orchestrator
    to: ResearchAgent
    protocol: async

skills_registry:
  - name: web.search
    type: http
  - name: db.query
    type: sql

---

Acceptance Criteria

• New openclaw generation mode
• Structured YAML/JSON output compatible with OpenClaw
• Automatic agent and skill inference from repository structure
• Documentation updated with usage example

[LorranHippolyte]

@vinilana fala tio! Tenho interesse em trabalhar nessa feature também. Poderia dar uma olhada e nos dizer se está liberado
para iniciarmos o desenvolvimento? Antes de começar, gostaríamos de alinhar a abordagem com você.

[matheuscbs]

Estou esperando pra participar e fazer o PR. Se pode seguir com essa config. Acho interessante alinharmos a abordagem e todo mundo poder contribuir um pouquinho.