intent-confirmation
当用户的请求在执行前需要澄清目标或边界时使用:需求抽象、涉及架构或设计决策、影响范围大、存在多种实现路径、可能修改重要文件,或用户明确要求先确认。不要用于简单问答、只读查询、明确的小修小改或可安全直接执行的任务。
Browse reusable Agent Skills, each with a clear purpose and practical guidance.
当用户的请求在执行前需要澄清目标或边界时使用:需求抽象、涉及架构或设计决策、影响范围大、存在多种实现路径、可能修改重要文件,或用户明确要求先确认。不要用于简单问答、只读查询、明确的小修小改或可安全直接执行的任务。
帮助用户把一个重复性任务设计成一个有边界的 Loop(运行契约 + 预算),并产出一份可复用的 loop 定义文档。 触发条件:(1) 用户想为某类反复执行的任务建立一个受控循环("每次都这样跑""反复打磨直到满足 X"), (2) 用户想为某个 R&K Flow Spec 定制阶段路径或修复循环预算, (3) 用户问"帮我写一个 loop / 设计一个循环"。 本 Skill 只负责交互式产出 loop 定义,不负责驱动执行。执行由对应的工作流 Skill 或用户自己承担。 不要用于一次性任务、简单问答,或已有明确 Spec 可直接执行的场景。
MUST READ before first MCP tool call in a session. Complete Envoy tool catalog with parameters and usage.
Use when you need to author, package, or share a skill folder — understanding the SKILL.md format, where skills live, and how to bundle a folder so it can be shared with others.
Upgrade the Nemotron Voice Agent to a new Pipecat (pipecat-ai) version. Reads release notes for every release in range, diffs old vs new, discovers every example pipeline and Pipecat call site, implements changes, then runs multi-agent gap analysis until clean. Generic across Pipecat versions.
Analyze the current conversation's tool calls, responses, errors, and approach to produce a structured feedback report for improving MCP tool usage with ARC-1. Use when the user asks to "analyze this session", "review my chat", "give feedback on tool usage", "how did I do", or wants a retrospective on a debugging session.
Use when a developer wants to add their own custom tool(s) to an ARC-1 MCP instance — an "extension" or "plugin" (FEAT-61). Guides the key architecture decisions (extension vs separate server; code tier vs manifest tier; which SAP API; scope/opType), then scaffolds the plugin and walks build + load + test. Do NOT use for adding a tool to ARC-1 core itself (that is an in-tree change), or for a different SAP backend (that is a separate server).
개발 중 발견한 중요한 지식을 데이터베이스에 능동적으로 축적하세요. 사용자가 요청하지 않아도, 버그 수정·패턴 발견·설계 결정·주의사항 등 가치 있는 지식이 발생하면 자동으로 memory-save를 호출하세요. 사용자가 "기억해둬", "저장해둬", "remember this" 등으로 명시할 때도 사용하세요.
실수에서 배우고 반복 패턴을 스킬로 자동 생성하세요. 사용자가 "아니", "잘못", "틀렸" 등을 말하면 실수를 기록하고 CLAUDE.md를 업데이트하세요. 같은 패턴이 3회 이상 반복되면 스킬로 승격하세요. mistake-log, skill-save MCP 도구를 능동적으로 사용하세요.