m3taversal
a6061ca968
- What: new divergence schema (structured disagreements between 2-5 claims), README rewrite with game framing, CLAUDE.md integration (knowledge structure, review checklist, quality gates) - Why: divergences are the core multiplayer mechanic — open questions that invite contributor evidence. Reviewed by Ganymede, Rhea, and Epimetheus across two rounds. Slimmed from 200 to 77 lines after over-engineering feedback. - Connections: unblocks seeding first divergences, enables importance-weighted scoring (coming soon) Pentagon-Agent: Leo <A3DC172B-F0A4-4408-9E3B-CF842616AAE1>
2.8 KiB
2.8 KiB
Divergence Schema
A divergence links 2-5 claims that offer competing answers to the same question. Not a bug — the most valuable part of the knowledge base. Every divergence is an open invitation: "We disagree about this — who's right?"
Where they live
domains/{domain}/divergence-{slug}.md — alongside the claims they reference. Cross-domain divergences go in the primary domain with secondary_domains.
YAML Frontmatter
---
type: divergence
title: "the question these claims disagree about"
domain: internet-finance | entertainment | health | ai-alignment | space-development | grand-strategy | mechanisms | living-capital | living-agents | teleohumanity | critical-systems | collective-intelligence | teleological-economics | cultural-dynamics
description: "why this disagreement matters and what resolving it would unlock"
status: open | resolved
claims: [] # 2-5 claim filenames
surfaced_by: "who identified this divergence"
created: YYYY-MM-DD
---
Body Format
# [question or tension]
[Why this matters. What changes if we knew the answer.]
## Divergent Claims
### [claim title]
**File:** [[claim-filename]]
**Core argument:** [1-2 sentences]
**Strongest evidence:** [what makes this credible]
### [claim title]
**File:** [[claim-filename]]
**Core argument:** [1-2 sentences]
**Strongest evidence:** [what makes this credible]
## What Would Resolve This
[Specific evidence contributors should look for. This is the research agenda — the game hook.]
## Cascade Impact
[What beliefs and positions change depending on which claim wins. This is the importance signal.]
---
Relevant Notes:
- [[related-claim]] — relationship
Topics:
- [[domain-map]]
Governance
- Who can propose: Any agent, any contributor, or surfaced during PR review
- Review: Leo reviews for genuine disagreement (not scope mismatch). Domain agents review claim summaries for accuracy.
- Resolution: Evidence-based only. No authority-based resolution.
When NOT to create a divergence
- Scope mismatch: Two claims about different scopes (structural vs functional, micro vs macro) aren't in tension. Fix the scope. ~85% of apparent tensions dissolve with better wording.
- Evidence gap: One claim simply lacks evidence. Strengthen or weaken the claim — don't create a divergence.
- False opposition: Complementary claims aren't contradictory. "AI helps diagnosis" and "AI doesn't help treatment" aren't in tension.
Divergences as game mechanic
Divergences are the highest-value contribution targets. Resolving one means:
- Changing claims in the KB
- Triggering cascade re-evaluation of beliefs and positions
- Demonstrating consequential knowledge
Importance-weighted contribution scoring is coming — the importance of a contribution will be proportional to the cascade impact of the divergence it helps resolve.