Scaffold and extend production Python projects for composite MCP servers (FastMCP), Strands agents with LTM/STM memory (AgentCore), hybrid pgvector search services, and OTEL/Phoenix tracing. Use when: starting new agentic projects, adding tools/namespaces/migrations, adding observability, or reviewing agentic stack projects.
npx claudepluginhub diegouis/provectus-marketplace --plugin proagent-agentic-stackThis skill uses the workspace's default tool permissions.
Scaffold and extend production agentic applications using Provectus's proven patterns.
Builds MCP servers in Python/TypeScript for AI agents, integrating APIs as tools with Pydantic/Zod schemas. Manages tool discovery, execution, and multi-server configs.
Scaffolds production-ready AI agents with Google's ADK: single/multi-agent orchestration, tool wiring, tests, and Vertex AI deployment.
Guides harness engineering for AI agents: context/memory management, guardrails, AGENTS.md/CLAUDE.md repo instructions, evals, observability, and orchestration.
Share bugs, ideas, or general feedback.
Scaffold and extend production agentic applications using Provectus's proven patterns.
DO NOT read reference files, run environment detection commands, or load mode files until the user has told you what they want to do.
MANDATORY: Call AskUserQuestion — do NOT render these as text:
AskUserQuestion( header: "Agentic Stack", question: "What do you need to build?", options: [ { label: "MCP Server", description: "Composite FastMCP server with tool mounting, DI, and ECS deployment" }, { label: "Strands Agent", description: "Agent with LTM/STM memory, AgentCore handler, and Bedrock deployment" }, { label: "Hybrid Search", description: "pgvector service with vector + full-text scoring and migrations" }, { label: "Add Tracing", description: "OTEL/Phoenix setup and custom span instrumentation" }, { label: "Extend Existing", description: "Add tools, memory namespaces, migrations, or Pulumi components" }, { label: "Review", description: "Audit an existing agentic stack project" } ] )
| User Intent | Mode File | Specialist Agent |
|---|---|---|
| New MCP server | commands/modes/scaffold-mcp-server.md | mcp-server-specialist |
| New Strands agent | commands/modes/scaffold-strands-agent.md | strands-agent-specialist |
| New hybrid search service | commands/modes/scaffold-hybrid-search.md | search-specialist |
| Add OTEL/Phoenix tracing | commands/modes/add-tracing.md | tracing-specialist |
| Add MCP tool to existing server | commands/modes/add-mcp-tool.md | mcp-server-specialist |
| Add memory namespace | commands/modes/add-memory-namespace.md | strands-agent-specialist |
| Add pgvector migration | commands/modes/add-migration.md | search-specialist |
| Add Pulumi component | commands/modes/add-pulumi-component.md | Domain-appropriate specialist |
| Review project | commands/modes/review.md | Domain-appropriate specialist |
CONTEXT GUARD: ONLY read when a mode file instructs you to.
| File | Contents |
|---|---|
references/shared-conventions.md | Pydantic settings, uv, Taskfile, Ruff/MyPy, Docker, SSM split |
references/mcp-server-patterns.md | FastMCP mount, DI, tool registration, lifespan |
references/strands-agent-patterns.md | Agent factory, memory, handler, session cache |
references/hybrid-search-patterns.md | Score formula, migrations, triggers, embeddings |
references/tracing-patterns.md | OTEL setup, custom spans, Phoenix, processors |
references/pulumi-patterns.md | Components, SSM loader, ECS/AgentCore stacks |