AI Engineering

Automatically trim, summarize, or drop stale content from the context window to save tokens while preserving task state. Use during long sessions, large file dumps, or before calling expensive models.

Switches the agent to a terse output mode, reducing token usage by ~50-70% while keeping code, commands, and technical terms exact. Use when you need brief replies or want to save costs.

A structured five-step thinking loop that helps AI agents make better decisions by weighing reversibility, gathering evidence, redesigning bets, attacking their own reasoning, and delivering a clear answer. Fires automatically on high-stakes choices or user indecision.

Use when creating, updating, splitting, merging, or restructuring an agent skill or skill library. Defines scoping rules and authoring conventions for deterministic behavior.

Analyzes local AI skill library inventory, usage, and test status in the terminal, surfacing candidates for deletion or downgrading.

Summarizes recent git commits and open PRs into a brief standup update with done, doing, and blockers bullets.

Spawn a panel of expert AI agents to debate any question using a structured protocol with iterative cross-examination and convergence detection, producing a decision landscape with confidence diagnostics.

Defines the Agent Skills open format, including directory structure, SKILL.md schema, naming conventions, and progressive disclosure model. Use this reference when creating, reviewing, or editing agent skills to follow the standard.

Organize durable agent research outputs as summaries, analysis, and evidence dossiers. Use when producing multi-layer research artifacts or coordinating research handoffs.

Build conversational multi-agent AI systems with Microsoft AutoGen. Covers AssistantAgent, UserProxyAgent, GroupChat, code execution, nested chats, tool integration, and MCP support. Use for conversation-driven multi-agent orchestration.

Synthesizes workflow bundles from accumulated state, generating structured output directories with umbrella skills, sub-skills, manifests, and decision maps for agent skill scanners.

Conducts a structured, adaptive interview to discover a user's workflow, mapping phases, tools, branching, and pain points to generate a workflow bundle.