Account for the Three-Tier Thinking Room (Paper VII §6.3) before lane work hardens

[flyingrobots]

Doctrine (AIΩN Paper VII §6.3, DOI 10.5281/zenodo.19751149)

Shared causal history creates exposure pressure a public/private split cannot
absorb. The protected working interval has three tiers:

1. Ephemeral Scratch — local, weakly retained, disposable.
2. Author-Only Speculative Lane — durable, replayable, sealed to the
   author by default.
3. Shared / Admitted Lane — collaboratively admitted causal truth.

Operational consequences the paper makes explicit: reading/replay/inspection
are revelation-only acts (never create causal truth); debugger-created
strands are real causal objects that default to Scratch or Author-Only
until an explicit promotion moves them into shared history; retained strand
provenance records creating principal, tool/session origin, fork basis, and
retention/revelation posture.

Why ASAP

Strand, braid, settlement, and session work is hardening now (#470, #476,
#483, #537, design 0025). If tier posture is not a field on lanes/strands
from the start, every one of those surfaces will assume "all lanes are
shared" and retrofitting author-sealing becomes a breaking migration.
James: "we should account for it ASAP rather than later."

Current state vs the tiers (@1465cfe7)

• Ephemeral Scratch exists by accident: StrandRegistry is in-memory and
  strands vanish on drop (DropReceipt only). No declared posture.
• Durable retention exists (WSC, retained evidence) with no author-seal or
  revelation tier — anything retained is implicitly shared-visible.
• 0025 sessions supply exactly the substrate tiers need (principal binding,
  session origin, lifecycle) but the design does not yet name tier posture.

Proposal sketch

• Add an explicit retention/revelation posture (tier) to strand creation
  and session-attributed work: scratch | author-only | shared.
• Default debugger/counterfactual forks to Scratch or Author-Only; promotion
  to Shared is an explicit, witnessed act (an Intent, per WARP_DRIFT's
  intent-only rule for topology mutation).
• Tie tier to the retention contract Λ in the shell family (Proposal: one hologram family across tick, braid, and suffix-transport scales (Paper VII Prop 3.5) #537): a shell's
  lawful revelation set depends on tier + observer rights.
• 0025 phase 2 should carry the tier field so sessions and tiers land
  together rather than as two migrations.

[coderabbitai]

🔗 Related PRs

#298 - feat(warp-core): TTD domain logic — provenance storage + causal cone walk [merged]
#326 - Echo contract hosting roadmap [merged]
#358 - feat(core): add admission evidence boundaries [merged]
#371 - feat(core): advance v0.1 contract evidence and retention [merged]
#380 - feat(core): expose retained evidence posture on readings [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!