Skill: bridge-based multi-hop reasoning

[Case 2]

• Initial Environment: A bounded article set about folklore, U.S. geography, and state capitals. The query links the canonical age needed for a Groundhog Day predictor to the state containing the U.S. capital, requiring multiple retrieved facts and an intermediate reasoning bridge.
• Real Question: How many years earlier would Punxsutawney Phil have to be canonically alive to have made a Groundhog Day prediction in the same state as the US capitol?
• Real Answer: A numeric answer derived by combining the predictor’s canonical start date with the state-capital bridge and a year-difference calculation.
• Why this demonstrates the capability: This case shows that multi-hop reasoning often combines bridging with a follow-on computation. The agent must follow the shared thread across articles before it can even set up the final calculation. The bridge is therefore the structural backbone of the reasoning path.

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