swe-debate-multi-agent-issue-resolution - Competitive Multi-Agent Debate for Issue Resolution
Orchestrates specialized agents to analyze software issues, debate fault hypotheses, and produce consolidated fix plans.
Tags
Updated: 2026-10-08Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Build code dependency graphs
- Trace fault propagation paths
- Generate ranked fault hypotheses
- Analyze issues from varied perspectives
- Orchestrate multi-round agent debates
- Consolidate root-cause and fix consensus
- Execute fixes through MCTS
Inputs
- Software repository
- Issue description
- Error location and message
- Code context
- LLM interface
Outputs
- Code dependency graph
- Fault propagation traces
- Ranked root-cause hypotheses
- Agent debate analyses
- Consolidated fix plan
- Applied code fixes
Requirements
- Python runtime
- Text-generation LLM
- Repository read access
- Repository write access for fixes