| name | proactive-conversational-engagement-synthesis |
| description | Use this skill when the user wants the agent to sound like a natural conversationalist who keeps the dialogue alive rather than providing dead-end answers. Trigger it for requests like 'make the agent more proactive', 'add follow-up questions to the results', 'simulate a back-and-forth chat', or 'encourage the user to keep asking things'. It is specifically geared towards generating synthetic training data where the assistant answers a query and then pivots to a related, inquisitive follow-up based on the retrieved knowledge. |
Skill: proactive-conversational-engagement-synthesis
1. Capability Definition & Real Case
- Professional Definition: The ability to generate responses that fulfill the immediate informational request while simultaneously analyzing the retrieved context to identify 'information scents'—adjacent, high-value facts that can be used to formulate proactive, contextually grounded follow-up questions. This ensures multi-turn dialogue continuity and simulates human-like curiosity, moving beyond static question-answering toward dynamic engagement.
- Dimension Hierarchy: Multi-Source Evidence Composition->Conversational Context Integration->proactive-conversational-engagement-synthesis
Real Case
[Case 1]
- Initial Environment: A technical report corpus regarding ISRO's (Indian Space Research Organisation) recent space docking mission (SPADEX). The text provides details about the mission's intent to demonstrate autonomous docking and its relevance to a future space station.
- Real Question: What is the primary goal of the SPADEX mission by ISRO?
- Real Trajectory: 1. Analyze the retrieved text for mission objectives. 2. Identify the core fact: demonstrating autonomous docking and refueling capability. 3. Scan for 'proactive hooks' in the text: mentioned connection to the future 'Bhartiya Antariksha Station'. 4. Formulate an answer containing the fact and a transition. 5. Generate a follow-up question regarding the space station to maintain engagement.
- Real Answer: ISRO's SPADEX mission is designed to demonstrate autonomous docking, rendezvous, and refueling capabilities between two spacecraft. Would you like to know how this technology will facilitate the assembly of the planned Bhartiya Antariksha Station?
- Why this demonstrates the capability: This case shows the transition from a factual retrieval task to proactive engagement. Instead of just stating the goal, the agent identifies the 'future implications' mentioned in the text and uses them to pose a follow-up, ensuring the conversation flows into the next logical topic (the space station).
[Case 2]
- Initial Environment: A regional culinary knowledge base describing traditional food and cafes in New Delhi, specifically mentioning 'rooftop dining' experiences and the cultural history of specific regions like Champa Gali.
- Real Question: Can you suggest a good cafe for coffee in South Delhi?
- Real Trajectory: 1. Search for coffee spots in South Delhi. 2. Retrieve information about 'Blue Tokai' in Champa Gali. 3. Observe secondary details in the context about the 'artistic vibe' and 'rooftop seating' available at that location. 4. Answer with the recommendation and then pivot to the specific rooftop experience as a hook. 5. Frame the follow-up in a natural, inquiring tone.
- Real Answer: You should definitely check out Blue Tokai in Champa Gali; it's known for its specialty roasts and unique artistic setting. Are you interested in cafes that also offer rooftop views for a sunset coffee experience?
- Why this demonstrates the capability: The agent demonstrates engagement by using sub-attributes of the retrieved entity (location and atmosphere) to suggest a new dimension of the query (rooftop dining), simulating how a human guide would provide a suggestion and then ask for preferences to narrow down the next step.
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:
-
Phase 1: Environment Exploration
Read the exploration guidelines to discover raw knowledge seeds:
references/EXPLORATION.md
-
Phase 2: Trajectory Selection
Once Phase 1 is complete, read the selection criteria to evaluate the trajectory:
references/SELECTION.md
-
Phase 3: Data Synthesis
Once a trajectory passes Phase 2, read the synthesis instructions to generate the final data:
references/SYNTHESIS.md
