Session Statistics: Every Inc AI-First Org Design Kit

Output Summary

Category Files Lines
AI-First Kit artifacts 41 9,208
Claude Code skills 5 284
CLAUDE.md 1 18
Research files 3 2,397
Grand Total 50 11,907

Artifacts by Category

Category Files Lines Description
Genome (identity) 7 618 Mission, values, voice, authority, tradeoffs, quality, anti-patterns
Governance 7 1,332 Authority matrix, boundaries, escalation, policy gen, ledger, learning loop, usage policy
Quality Gates 9 386 4 gates + 4 holdout sets + index (23 holdout scenarios total)
Specifications 4 656 Article production, compound engineering, consulting, podcast
Roles 1 553 14 roles decomposed with Three-Variable Model
Agents 5 1,021 5 agent configs + index
Adoption 2 973 Maturity ladder (19 assessments) + 2 sprint designs
Evolution 2 524 Baseline audit + decision ledger
Operational 2 2,846 AGENT-PRIMER (175 lines) + ORG-DESIGN-DUMP (2,671 lines)
Coordination audit 1 149 Initial diagnostic
Political map 1 150 Stakeholder analysis with reframes

Research Phase

Source Articles Fetched Insights Extracted
Chain of Thought (Dan Shipper) 20 542 lines
Source Code (engineering) 20 986 lines
On Every (company updates) 23 869 lines
Total 63 2,397 lines

Additional web searches: ~30 queries across multiple agents

Agent Utilization

Phase Agents Spawned Parallel? Purpose
Research wave 1 3 Yes Company research, CEO research, plugin analysis
Research wave 2 2 Yes Key articles, consulting/products pages
Research wave 3 3 Yes Chain of Thought, Source Code, On Every article batches
Phases 5-6 2 Yes Specifications + roles
Phases 7+9 2 Yes Governance + maturity ladder
Phases 8+10+11 3 Yes Operationalize + sprints + usage policy
Phases 12+13 2 Yes Agent builder + evolution auditor
Quality audit 2 Yes Artifact consistency check + statistics
Total agents 19    

Skills Invoked

Skill Phase Invocation Method
coordination-audit 1 Direct (main agent)
org-genome-builder 2 Direct (main agent)
political-navigator 3 Direct (main agent)
quality-gate-designer 4 Direct (main agent)
specification-writer 5 Delegated to subagent
role-value-mapper 6 Delegated to subagent
governance-architect 7 Delegated to subagent
operationalize 8 Delegated to subagent (initial) + Direct (re-run)
maturity-ladder 9 Delegated to subagent
adoption-sprint-designer 10 Delegated to subagent
usage-policy-writer 11 Delegated to subagent
agent-builder 12 Delegated to subagent
evolution-auditor 13 Delegated to subagent

Key Metrics

  • Compression ratio: 6,868 lines of source → 206 lines of AGENT-PRIMER (33:1) [post-revision; original session: 9,208 → 175 = 52:1]
  • Holdout scenarios: 23 across 4 gates
  • Hard boundaries: 9 non-negotiable rules
  • Team members assessed: 19 (maturity ladder)
  • Roles mapped: 14 (Three-Variable Model)
  • Agent configurations: 5 (editorial, engineering, consulting, distribution, product GM) [product GM added in revision]
  • Claude Code skills generated: 5 (/org-* commands)
  • Value conflicts modeled: 5 specific tradeoff rules
  • Anti-patterns documented: 10 organizational taste markers