Tier 4

agi - ARAW to GOSM Integration

ARAW to GOSM Integration

Overview

Systematic procedure for translating ARAW exploration outputs into GOSM execution inputs.

ARAW = Exploration engine (maps possibility space) GOSM = Execution engine (turns insights into action)

This bridge ensures ARAW findings automatically populate GOSM structures.

Steps

Step 1: Run ARAW synthesis

If multiple ARAW databases exist: python synthesize.py *.db —find-tensions —extract-actions actions.json

This produces:

  • Summary of each database
  • Cross-database themes
  • Tensions and contradictions
  • Consolidated action items (DO_FIRST, DO_SECOND, crux nodes)

Step 2: Extract MUST_VERIFY items

From synthesis output, identify:

MUST_VERIFY (before strategy commitment):

  • All crux nodes
  • All DO_FIRST actions
  • All unresolved tensions

These become GOSM gates that must pass before proceeding.

Step 3: Populate Assumption Register

For each crux node:

assumption: id: [from ARAW] claim: [original claim text] confidence: [ARAW probability estimate] criticality: CRITICAL | HIGH | MEDIUM | LOW validation_method: [how to test] validation_status: UNVALIDATED if_false_action: [contingency] source: ARAW:[database_name]

Step 4: Populate Risk Register

For each claim with probability < 0.7:

risk: id: [generated] description: [claim might be false] probability: [1 - ARAW probability] impact: [from leverage score] mitigation: [if applicable] contingency: [ARAW ASSUME_WRONG branch] source: ARAW:[database_name]

Step 5: Identify Strategy Candidates

From ARAW branches, extract:

  • High-leverage ASSUME_RIGHT branches = Strategy candidates
  • High-leverage ASSUME_WRONG branches = Alternative approaches

Feed into Strategy Discovery phase.

Step 6: Create Decision Tree Branches

For each ARAW tension:

decision_point: id: [generated] question: [tension summary] option_a: [one side of tension] option_b: [other side of tension] resolution_criteria: [how to decide] current_choice: UNRESOLVED source: ARAW_TENSION

Step 7: Mark High-Confidence Foundations

For themes appearing in 2+ ARAW runs:

foundation: claim: [theme summary] confidence: HIGH (cross-verified) sources: [list of ARAW databases] validation_needed: MINIMAL

These can be treated as near-axioms for planning.

When to Use

  • After ARAW exploration on a problem/goal
  • Before starting GOSM planning phase
  • When ARAW identifies tensions or crux nodes