AI Engineering

Agentic Actions Auditor provides static security analysis guidance for GitHub Actions workflows that invoke AI coding agents, detecting attack vectors where attacker-controlled input reaches an AI agent. Use it when auditing CI/CD pipeline security for prompt injection risks or evaluating agentic action configurations.

Agent Orchestration Advisor helps product managers design multi-agent AI workflows with clear boundaries, handoffs, and monitoring. Use it when a complex PM task should run as parallel specialized agents instead of one linear process.

AI-Shaped Readiness Advisor assesses whether your product work is AI-first or AI-shaped by evaluating five essential PM competencies and recommending which capability to build next. Use it to gauge AI maturity and identify gaps in how your team operates with AI.

Altitude Horizon Framework defines the two-axis mental model (altitude and horizon) that distinguishes Director-level thinking from PM thinking. Use it to diagnose scope, time-horizon, or leadership-level gaps in the PM-to-Director transition.

Autonomous Investigation defines the canonical contract for investigation skills: AI research that proceeds without the user, using question budgets, search-plan gates, evidence labels, confidence stacking, and stable diffable outputs. Use it when delegating research to agents or scheduled runs.

Context Engineering Advisor helps product managers diagnose context stuffing versus context engineering and apply bounded contexts, episodic retrieval, and the Research→Plan→Reset→Implement cycle to reduce AI workflow bloat.

Director Readiness Advisor guides product managers through the PM-to-Director transition across preparing, interviewing, landing, and recalibrating, offering adaptive questions and targeted coaching. Use it when leadership scope is changing and you need practical, situation-specific advice.

Add Provider integrates new DeepChat LLM providers by generating explicit source changes across provider defaults, registry, transports, and tests. Use when a developer asks Codex to add a provider, provider profile, upstream provider config, model catalog mapping, or a special provider adapter.

DeepChat CLI controls the running DeepChat desktop application through its bundled command-line interface for model inference, image/video/speech generation, transcription, OCR, artifact inspection, public configuration, Skills, and MCP operations.

MCP Builder guides the creation of high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use it when building MCP servers to integrate external APIs, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

Memory Management guides an agent to recall prior work, persist durable conclusions, and route reusable procedures or recurring needs into Skills, Scheduled Tasks, or Tape. Use it when a task may produce durable learning or when the user asks to recall, remember, or continue earlier work.

Skill Creator guides users through creating or updating skills that extend Claude's capabilities with specialized knowledge, workflows, and tool integrations. Use it when building a new skill or refining an existing one.