Skip to main content

/think — Deep Reasoning Mode

The /think workflow activates maximum reasoning depth, executing all analysis phases without shortcuts. Use this for high-stakes decisions, complex problems, and when you need exhaustive analysis.

When to Use

  • High-stakes decisions (>$10K or >6 months impact)
  • Complex multi-variable problems
  • Deep psychological processing
  • When user explicitly wants exhaustive analysis
  • When destruction risk is suspected

Behavior

When /think is invoked, Athena applies rigorous analysis:
  1. All Phases Mandatory - Executes Phase 0 through Phase VII
  2. Counterfactual Required - Phase 3F (Upward/Lateral/Downward) always included
  3. Confrontation Required - Phase VII always included
  4. Multi-Path Reasoning - 2-3 branches minimum, including dead ends
  5. Stress-Test - Phase 5E robustness check mandatory
  6. No Shortcuts - Even if query seems simple, go full depth

Phase Sequence (Mandatory)

Every /think invocation follows this complete sequence:
  • Phase 0: Graph of Thought
  • Phase I: Upstream Tracing (L1→L5)
  • Phase II: Validation
  • Phase III: Analysis (3A-3F)
  • Phase IV: Probabilistic Modeling
  • Phase V: Strategic Advice (5A-5E)
  • Phase VI: Final Thoughts
  • Phase VII: Confrontation

Output Format

Full Tri-Brid template with all phases. Expect longer latency and higher token usage (~2000 tokens).

Usage Example

User: /think Should I leave my job to pursue trading full-time?

Athena: [Executes all phases, 2000+ word analysis]

Phase 0: Graph of Thought
→ Central question: Career transition under financial uncertainty
→ Upstream: Current job satisfaction, runway, skill readiness
→ Downstream: Income stability, identity shift, reversibility

Phase I: Upstream Tracing
- L1: Immediate trigger (frustration with current role)
- L2: Financial foundation (6-month runway vs 12-month ideal)
- L3: Trading skill level (profitable for 8 months, sample size?)
- L4: Identity: How much of self-worth is tied to "trader" label?
- L5: Life context: Dependents? Health insurance? Age/career stage?

[... continues through all phases ...]

Phase VII: Confrontation
The harsh truth: You're asking permission because you already doubt.
If the data was overwhelmingly positive, you'd have quit already.
The real question isn't "should I?" — it's "what's holding me back?"
Answer that first.

Use Cases

  • Career transitions and major life changes
  • Large financial decisions (investments, purchases)
  • Relationship decisions with long-term impact
  • Technical architecture decisions
  • Strategic business pivots

Comparison with Other Reasoning Modes

ModeDepthToken BudgetUse Case
NormalLow~200 tokensSimple queries, quick answers
/thinkHigh~2000 tokensImportant decisions, complex problems
/ultrathinkMaximum~5000+ tokensLife-altering decisions, multi-stakeholder analysis

Rule of Thumb

Default to normal mode. Escalate to /think for $10K+ decisions. Use /ultrathink only for maximum-depth analysis.

Build docs developers (and LLMs) love