Skip to main content

interactive-information-gathering

Use this skill when the user’s request sounds actionable but is still missing a key detail, so the agent has to ask smart follow-up questions instead of guessing. Trigger it for situations like returns, cancellations, bookings, or account changes where one missing detail changes the correct action. Everyday examples include: 'I want to change my order', 'can you cancel my flight', 'help me return this item', and 'fix this booking but only if the cheaper option works.'

설치로 이동

소스 정보

저장소
Dingxingdi/paper_fast_search_backup
최근 소스 활동
2026년 4월 8일 15:14
감지된 SKILL.md 언어
영어
스타
0
포크
0

설치 방법

기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.

소스 파일 검토

설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.

파일 탐색기
4 개 파일

SKILL.md 표시 중

SKILL.md
소스 지침 · 읽기 전용 미리보기
name
interactive-information-gathering
description
Use this skill when the user’s request sounds actionable but is still missing a key detail, so the agent has to ask smart follow-up questions instead of guessing. Trigger it for situations like returns, cancellations, bookings, or account changes where one missing detail changes the correct action. Everyday examples include: 'I want to change my order', 'can you cancel my flight', 'help me return this item', and 'fix this booking but only if the cheaper option works.'
# Skill: interactive-information-gathering ## 1. Capability Definition & Real Case * **Professional Definition**: The ability to acquire missing task parameters through natural dialogue, asking for the right information at the right time before taking irreversible actions or giving final recommendations. * **Dimension Hierarchy**: Conversational Effectiveness->Task-oriented Dialogue Control->interactive-information-gathering ### Real Case **[Case 1]** * **Initial Environment**: A customer-service chat begins with a user asking to return one product and exchange two others. The agent can read a retail order database through tools, but several details such as payment method preference and exact eligible items still need confirmation before any write action. * **Real Question**: I want to return the water bottle, and exchange the pet bed and office chair to the cheapest version. Mention the two things together. * **Real Trajectory**: 1. Read the relevant order to identify item IDs, order state, and eligible operations. 2. Ask any missing clarification needed to make the return and exchange valid under policy. 3. Explain the available outcomes and money implications before acting. 4. Perform only the confirmed database write that matches the user’s stated preference. * **Real Answer**: A response that gathers the missing operational details, explains the feasible path, and completes only the confirmed return-or-exchange action sequence. * **Why this demonstrates the capability**: The user intent is partially specified and spans multiple operations. The assistant must not guess hidden parameters or act prematurely; it must gather information interactively while keeping the task coherent and policy-compliant. ## Pipeline Execution Instructions To synthesize data for this capability, you must strictly follow a 3-phase pipeline. **Do not hallucinate steps.** Read the corresponding reference file for each phase sequentially: 1. **Phase 1: Environment Exploration** Read the exploration guidelines to discover raw knowledge seeds: `references/EXPLORATION.md` 2. **Phase 2: Trajectory Selection** Once Phase 1 is complete, read the selection criteria to evaluate the trajectory: `references/SELECTION.md` 3. **Phase 3: Data Synthesis** Once a trajectory passes Phase 2, read the synthesis instructions to generate the final data: `references/SYNTHESIS.md`
GitHub에서 보기