菜单
Multi-subagent framework. Allows the main agent to spawn specialized Claude Code subagents with specific skills. Each subagent runs autonomously (full agentic loop) and returns a structured result. Engine: Claude Code CLI (claude --bare -p). Trigger: "subagent", "spawn agent".
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Automated research subagent. Accepts a natural-language query, gathers sources from the knowledge graph (KG + BM25 + E5 + Web + Notion), performs SGR synthesis, generates validated knowledge cards, and writes them to the graph. Explicit trigger: "autosearch" — activates the subagent without confirmation. Soft triggers (require confirmation): "research", "analyze", "compare", "deep dive", "review".
SGR-guided knowledge graph node creation with canonical schema validation. Use when creating new nodes, updating existing nodes, or auditing graph health. Enforces closed vocabularies for node types, clusters, and relation types to prevent taxonomy drift. Triggers on: "create node", "add to graph", "new insight", "graph audit", "validate graph", "taxonomy check".
SGR-enhanced pool of cheap specialized subagents based on sgr-agent-core. Uses Schema-Guided Reasoning to boost mini/Flash models to near-Sonnet quality. Trigger: "sgr", "sgr pool", "cheap agent", "mini agent", "sgr-researcher", "sgr-verifier", "sgr-summarizer".
Safe OpenClaw gateway configuration changes with validation and auto-rollback
Design-before-code protocol. Routes new features, refactors, and complex changes through the appropriate design artifact BEFORE implementation begins. Use when the user asks to design a feature, write an RFC, think through architecture, plan a new skill, structure a large refactor, or define the implementation approach before coding. Triggers: "design", "RFC", "architecture", "new skill", "new feature", "large change", "need a plan", "design-architect", "tech spec", "ADR".
Subagent for generating PPTX presentations via Claude Code CLI + official pptx skill. Trigger: "presentation", "slides", "deck".
| name | subagent |
| description | Multi-subagent framework. Allows the main agent to spawn specialized Claude Code subagents with specific skills. Each subagent runs autonomously (full agentic loop) and returns a structured result. Engine: Claude Code CLI (claude --bare -p). Trigger: "subagent", "spawn agent". |
When a task requires a specialized agent with specific skills:
Main Agent → SubagentExecutor.prepare() → SubagentRequest (CLI command)
→ SubagentExecutor.execute() → SubagentResult
Engine: claude --bare -p with --add-dir for skills, --output-format json.
from skills.subagent.executor import SubagentExecutor
executor = SubagentExecutor() # local mode
# Or remote (VPS)
executor = SubagentExecutor(mode="remote", host="root@YOUR_VPS_IP")
# Two steps (analogous to AutoResearch plan→gather)
request = executor.prepare("presentation-builder", "Create a presentation...")
result = executor.execute(request)
# Or as a single call
result = executor.run("presentation-builder", "Create a presentation...")
print(result.summary())
Registry: 99_process/subagent_registry.yaml (field executor for CLI-based agents).
| Agent ID | Skills | Description |
|---|---|---|
presentation-builder | pptx | PPTX presentation generation |
skills/<name>/api.py with a <Name>Agent class99_process/subagent_registry.yaml with the executor fieldconfig/agent_scopes.jsonSubagentExecutor will pick it up automatically| Mode | How | When |
|---|---|---|
local | subprocess.run(claude ...) | Claude Code CLI installed locally |
remote | ssh host "claude ..." | Claude Code CLI on VPS |
Мультисубагентний фреймворк для спавну спеціалізованих Claude Code субагентів з конкретними скілами. Кожен субагент працює автономно та повертає структурований результат. Підтримує локальний та віддалений (VPS) режими.