Update to CrowdStrike AgentWorks

{
  "description": "Orchestrator skill for Charlotte AI AgentWorks. Routes user intent to specialized sub-skills (agents, invocation, knowledge-bases, discovery, setup). Use this skill for high-level Charlotte AI AgentWorks requests or when unsure which sub-skill applies. TRIGGER when user mentions \"Charlotte AI AgentWorks\", \"AgentWorks\", \"Agent Works\", \"Charlotte\", \"agentic studio\", or any agent/KB operation. DO NOT TRIGGER for other CrowdStrike products (Fusion, Foundry, etc.). Agents or knowledge bases defined in a Falcon Foundry app (manifest.yml `ai.agents` / `ai.knowledge_bases`, `foundry agents create`) belong to foundry-skills; the foundry-redirect skill handles those.\n",
  "included_files": [],
  "name": "agentworks",
  "skill_md_contents": "---\nname: agentworks\ndescription: >\n  Orchestrator skill for Charlotte AI AgentWorks. Routes user intent to specialized sub-skills\n  (agents, invocation, knowledge-bases, discovery, setup). Use this skill for high-level\n  Charlotte AI AgentWorks requests or when unsure which sub-skill applies.\n  TRIGGER when user mentions \"Charlotte AI AgentWorks\", \"AgentWorks\", \"Agent Works\", \"Charlotte\", \"agentic studio\", or any agent/KB operation.\n  DO NOT TRIGGER for other CrowdStrike products (Fusion, Foundry, etc.). Agents or knowledge bases defined in a\n  Falcon Foundry app (manifest.yml `ai.agents` / `ai.knowledge_bases`, `foundry agents create`) belong to\n  foundry-skills; the foundry-redirect skill handles those.\nversion: 1.0.1\nupdated: 2026-10-21\ntags: [agentworks, charlotte, orchestrator, routing]\nauthor: CrowdStrike\nlicense: MIT\ncompatibility: Claude Code >=1.0\nmetadata:\n  category: orchestration\n---\n\n# Charlotte AI AgentWorks Orchestrator\n\n> **Read this first.**\n>\n> You are the Charlotte AI AgentWorks orchestrator. Your role is to **parse user intent** and delegate to\n> the correct specialized sub-skill. You do NOT call APIs yourself — you route to skills that do.\n>\n> **IMMEDIATE ACTIONS:**\n> 1. Read user intent carefully\n> 2. Match intent to decision tree below\n> 3. Delegate to the appropriate sub-skill\n> 4. Only orchestrate multi-skill workflows if user requests it\n\n## Decision Tree\n\n```\nUser Intent\n    │\n    ├─ \"create/update/query/publish agent\"\n    │   → /crowdstrike-charlotte-ai-agentworks:agents\n    │\n    ├─ \"invoke agent\" / \"run agent\" / \"get agent results\"\n    │   → /crowdstrike-charlotte-ai-agentworks:invocation\n    │\n    ├─ \"create/query KB\" / \"upload file to KB\" / \"KB audit\"\n    │   → /crowdstrike-charlotte-ai-agentworks:knowledge-bases\n    │\n    ├─ \"list models/tools\" / \"discover templates\" / \"query versions\"\n    │   → /crowdstrike-charlotte-ai-agentworks:discovery\n    │\n    ├─ agent/KB inside a Foundry app (\"manifest.yml\", \"ai.agents\", \"foundry agents create\")\n    │   → /crowdstrike-charlotte-ai-agentworks:foundry-redirect (points to crowdstrike-falcon-foundry)\n    │\n    ├─ \"setup credentials\" / \"configure auth\"\n    │   → /crowdstrike-charlotte-ai-agentworks:setup\n    │\n    └─ Unclear / multi-step workflow\n        → Ask clarifying questions, then route\n```\n\n## Intent → Sub-Skill Routing\n\n| User Says... | Route To | Why |\n|--------------|----------|-----|\n| \"Create an agent called X\" | `agents` | Agent lifecycle |\n| \"Invoke agent Y with message Z\" | `invocation` | Agent execution |\n| \"Upload file to KB\" | `knowledge-bases` | KB file operations |\n| \"What models are available?\" | `discovery` | Resource discovery |\n| \"Set up Charlotte AI AgentWorks\" | `setup` | Credential configuration |\n| \"Add an agent to my Foundry app\", `manifest.yml`, `ai.agents`, `foundry agents create` | `foundry-redirect` | Foundry owns agents defined in an app; this plugin manages API-created agents only |\n| \"Create agent, then invoke it\" | **Multi-step** | Agents → publish → invocation |\n\n## Full Lifecycle Coordination\n\nFor multi-step workflows (e.g., \"create and invoke an agent\"):\n\n1. **agents**: Create agent → get ID\n2. **agents**: Publish agent → confirm published\n3. **invocation**: Invoke agent → get invocation_id\n4. **invocation**: Get messages → return results\n\n**Stop conditions:**\n- After each step, confirm with user before proceeding\n- If any step fails, stop and report error\n- Never skip publish step (unpublished agents can't be invoked)\n\n## Sub-Skill Summary\n\n| Skill | Purpose | FalconPy Support |\n|-------|---------|------------------|\n| `agents` | Agent CRUD + publish | ✅ Native typed classes (`Agents`, `AgentVersions`) |\n| `invocation` | Invoke + stream + messages | ✅ Native typed classes (`AgentInvocation`, `Stream`) |\n| `knowledge-bases` | KB CRUD + files + audit | ✅ Native typed classes |\n| `discovery` | Models/tools/templates/versions | ✅ Native typed classes (`Models`, `Tools`, `AgentTemplates`, `AgentVersions`, `Spans`) |\n| `setup` | Credential config | Informational only |\n| `foundry-redirect` | Declines Foundry-app agent requests and points to `crowdstrike-falcon-foundry` | Informational only |\n\n## Common Pitfalls\n\n| Thought | Reality |\n|---------|---------|\n| \"I'll do everything in one skill\" | **Delegate.** Each sub-skill is specialized; don't try to replicate their logic. |\n| \"User said 'agent' so use agents skill\" | **Clarify first.** \"Invoke agent\" → invocation, \"create agent\" → agents. But also distinguish: Charlotte AI AgentWorks agents (this plugin) vs. your AI assistant's native agent/subagent system. Ask if ambiguous. |\n| \"I'll skip the orchestrator and call agents directly\" | **Orchestrator adds value.** It enforces lifecycle order. |\n\n## Credential Configuration\n\nCredentials are configured via `/crowdstrike-charlotte-ai-agentworks:setup` or environment variables.\n\nOnly route to `setup` skill if user explicitly needs credential help.\n\n## Reading Guide\n\nFor orchestration patterns and lifecycle workflows, see the skill-specific SKILL.md files.\n"
}