| name | agi-reasoning-engine |
| description | Advanced reasoning and problem-solving framework for complex, multi-step tasks. Use when OrcBot needs to perform deep planning, cross-domain knowledge synthesis, or high-reliability execution with self-correction. |
AGI Reasoning Engine
This skill elevates OrcBot's reasoning capabilities by enforcing a structured approach to complex problem-solving. It prioritizes first-principles thinking, rigorous verification, and adaptive strategy.
Core Directives
1. Multi-Dimensional Planning
For any non-trivial task, do not jump to execution. Instead:
- Deconstruct: Break the objective into atomic sub-tasks.
- Identify Dependencies: Map the logical flow and critical paths.
- Edge Case Analysis: Proactively identify potential failure modes.
2. First-Principles Thinking
When faced with a roadblock:
- Strip away assumptions.
- Re-examine the fundamental constraints.
- Rebuild the solution from the most basic truths known about the system.
3. Iterative Verification Loop
Every significant action must follow the Act -> Verify -> Refine cycle:
- Act: Perform the operation.
- Verify: Use empirical evidence (tests, logs, outputs) to confirm success.
- Refine: If the result deviates from the expected outcome, diagnose the root cause and adjust the strategy.
4. Cross-Domain Synthesis
Leverage knowledge across multiple technical domains (e.g., combining systems architecture with security best practices) to ensure holistic solutions.
Reasoning Patterns
Refer to references/reasoning-patterns.md for specific mental models and heuristics to apply during execution.
Triggering Self-Correction
If you detect a contradiction or a logical gap in your current approach, stop immediately. State the contradiction clearly and re-evaluate your strategy using the First-Principles directive.
