BaseInfinity · BaseInfinity · Jul 10, 2026 · Jul 11, 2026
diff --git a/AGENTS.md b/AGENTS.md
@@ -13,7 +13,7 @@ Meta-repository — SDLC Wizard documentation, automation, and a zero-dep Node.j
 
 ## AI Setup Lanes
 
-This repo recommends four setup lanes — Setup A (Sonnet 5 + Fable advisor, GPT-5.5 xhigh reviewer, recommended default), Setup B (Opus 4.6 Stability: Fable advisor, Opus 4.6 max driver, GPT-5.5 xhigh reviewer — legacy flagship), Setup C (OpusPlan Hybrid: Opus 4.8 plan mode + Sonnet 5 execute, GPT-5.5 xhigh reviewer), and Setup D (Claude Lite: Sonnet driver, no reviewer). See [`AI_SETUP_LANES.md`](AI_SETUP_LANES.md) for the full pick list.
+This repo recommends four setup lanes — Setup A (Sonnet 5 + Fable advisor, GPT-5.6 Sol xhigh reviewer, recommended default), Setup B (Opus 4.6 Stability: Fable advisor, Opus 4.6 max driver, GPT-5.6 Sol xhigh reviewer — legacy flagship), Setup C (OpusPlan Hybrid: Opus 4.8 plan mode + Sonnet 5 execute, GPT-5.6 Sol xhigh reviewer), and Setup D (Claude Lite: Sonnet driver, no reviewer). See [`AI_SETUP_LANES.md`](AI_SETUP_LANES.md) for the full pick list.
 
 The lanes are guidance, not a hard rule — maintainer override is always allowed.
 

diff --git a/AI_SETUP_LANES.md b/AI_SETUP_LANES.md
@@ -10,10 +10,10 @@ This is **guidance, not a hard rule**. Maintainer override is always allowed.
 |------|-------|--------|
 | **Advisor** | Fable 5 (via `advisorModel: "fable"`) | `high` (server-side) |
 | **Driver** | Sonnet 5 (`claude-sonnet-5`) | `high` default, `/effort xhigh` for hard tasks |
-| **Reviewer** | Codex (GPT-5.5) xhigh | — |
+| **Reviewer** | Codex (GPT-5.6 Sol) xhigh | — |
 | **Escalation** | Opus 4.8 xhigh or Fable 5 review | When stuck or high-stakes |
 
-The new standard. Sonnet 5 beats Opus 4.6 on every coding benchmark (SWE-bench Verified 85.2% vs 80.8%, Terminal-Bench 80.4% vs 65.4%) while using ~5x less Max quota per turn. Fable 5 advises at key decision points via native `advisorModel` (v2.1.170+). GPT-5.5 xhigh reviews cross-family. Escalate to Opus 4.8 xhigh for the hardest debugging or architecture decisions — don't run Opus as the daily driver (burns Max limits 2-3x faster).
+The new standard. Sonnet 5 beats Opus 4.6 on every coding benchmark (SWE-bench Verified 85.2% vs 80.8%, Terminal-Bench 80.4% vs 65.4%) while using ~5x less Max quota per turn. Fable 5 advises at key decision points via native `advisorModel` (v2.1.170+). GPT-5.6 Sol xhigh reviews cross-family. Escalate to Opus 4.8 xhigh for the hardest debugging or architecture decisions — don't run Opus as the daily driver (burns Max limits 2-3x faster).
 
 **Effort escalation ladder:** Start at `high` (Sonnet 5's default and sweet spot). Raise to `xhigh` when doing hard debugging, multi-file migrations, or long agent runs. `max` is rarely worth it — doubles cost for marginal gains per CodeRabbit testing.
 
@@ -25,7 +25,7 @@ The new standard. Sonnet 5 beats Opus 4.6 on every coding benchmark (SWE-bench V
 |------|-------|--------|
 | **Advisor** | Fable 5 (via `advisorModel: "fable"`) | `high` (server-side) |
 | **Driver** | Opus 4.6 (`claude-opus-4-6`) | `max` |
-| **Reviewer** | Codex (GPT-5.5) xhigh | — |
+| **Reviewer** | Codex (GPT-5.6 Sol) xhigh | — |
 
 The consistency-first lane. Opus 4.6 is the only model where `max` effort works without overthinking — months of field data, Active through at least Feb 2027, proven predictable. Lower benchmark scores than Sonnet 5 but higher consistency. Choose this when you've tuned prompts to Opus 4.6 behavior and reliability matters more than peak capability.
 
@@ -38,9 +38,9 @@ The consistency-first lane. Opus 4.6 is the only model where `max` effort works
 | **Planner** | Opus 4.8 (via Plan Mode — Shift+Tab) | `xhigh` |
 | **Advisor** | Fable 5 or Opus 4.8 (via `advisorModel`) | — |
 | **Driver** | Sonnet 5 (auto execute mode) | `high` |
-| **Reviewer** | Codex (GPT-5.5) xhigh | — |
+| **Reviewer** | Codex (GPT-5.6 Sol) xhigh | — |
 
-Cost-efficient hybrid using CC's native `opusplan` alias. Opus 4.8 reasons during Plan Mode, Sonnet 5 executes. Max-bundled — no API credit drain. Pin `model: "opusplan"` + `advisorModel: "fable"` in project settings. Sonnet 5 now uses 1M context natively (no `[1m]` suffix needed). GPT-5.5 xhigh is the cross-model reviewer.
+Cost-efficient hybrid using CC's native `opusplan` alias. Opus 4.8 reasons during Plan Mode, Sonnet 5 executes. Max-bundled — no API credit drain. Pin `model: "opusplan"` + `advisorModel: "fable"` in project settings. Sonnet 5 now uses 1M context natively (no `[1m]` suffix needed). GPT-5.6 Sol xhigh is the cross-model reviewer.
 
 **Note:** Opus 4.6 cannot advise Sonnet 5 (rejected in the advisor pairing table). Use Fable 5 or Opus 4.8 as advisor for this lane.
 
@@ -116,11 +116,13 @@ Setup D is for work where SDLC discipline overhead exceeds the value:
 
 ## Final Review Policy
 
-**Setups A, B, and C end at GPT-5.5 xhigh as the cross-model reviewer.** Claude can't grade its own homework — the reviewer always belongs to a different lab with different blind spots. See [CLAUDE_CODE_SDLC_WIZARD.md → "Cross-Model Review (Codex)"](CLAUDE_CODE_SDLC_WIZARD.md) for the handoff protocol.
+**Setups A, B, and C end at GPT-5.6 Sol xhigh as the cross-model reviewer.** Claude can't grade its own homework — the reviewer always belongs to a different lab with different blind spots. See [CLAUDE_CODE_SDLC_WIZARD.md → "Cross-Model Review (Codex)"](CLAUDE_CODE_SDLC_WIZARD.md) for the handoff protocol.
 
 **Setup D has no reviewer** — the blast radius doesn't justify it. If you're unsure whether a task is truly Lite, it probably isn't. Escalate.
 
-If GPT-5.5 isn't available on your OpenAI account, Codex auto-falls back to GPT-5.4 — still keep `model_reasoning_effort="xhigh"`. Lower reasoning misses subtle bugs that the reviewer is the last gate to catch.
+If GPT-5.6 Sol isn't available on your OpenAI account, Codex auto-falls back to Terra — still keep `model_reasoning_effort="xhigh"`. Lower reasoning misses subtle bugs that the reviewer is the last gate to catch.
+
+**Escalation for unusually risky PRs:** `xhigh` is the evidence-based default — OpenAI's own migration guidance is to preserve the prior effort baseline, and no published data shows `max` or Pro mode catching meaningfully more real bugs than `xhigh` on ordinary PR review. For a PR you'd genuinely lose sleep over (security-sensitive, high blast radius, touches the installer or a consumer-facing template), escalate the reviewer to `max` or Pro mode — a once-per-PR gate is exactly the kind of low-frequency, high-stakes call site where the extra cost is easiest to justify. Don't make it the default.
 
 ## Version Requirement
 
@@ -159,11 +161,11 @@ Whichever path you use, the cross-model PR review gate still applies.
 
 - Drop to Setup A (Sonnet 5) for the remainder of the day — same discipline, far less quota burn
 - Or drop to Setup D for grunt work that doesn't need deep reasoning
-- Or use Sonnet directly for the final mechanical edits, then run the GPT-5.5 reviewer over the whole diff at the end
+- Or use Sonnet directly for the final mechanical edits, then run the GPT-5.6 Sol reviewer over the whole diff at the end
 
 **Setup D uses Sonnet** — same model as Setup A's driver, Max-bundled. One less model to manage if you're already on Setup A.
 
-The reviewer (GPT-5.5 xhigh) is billed against your OpenAI account, separately. Watch both bills.
+The reviewer (GPT-5.6 Sol xhigh) is billed against your OpenAI account, separately. Watch both bills.
 
 ## Autocompact Thresholds
 
@@ -201,11 +203,11 @@ Credit allocations: Pro $20/mo, Max 5x $100/mo, Max 20x $200/mo. **No rollover.*
 
 ### What this means for the lanes
 
-- **Setup A — Sonnet 5 + Fable advisor:** Sonnet 5's native 1M context — interactive session, Max-bundled, no `[1m]` suffix needed. Fable 5 advisor via `advisorModel: "fable"` — also Max-bundled. GPT-5.5 xhigh reviewer on ChatGPT subscription. Roughly 5x less Max quota consumed per turn than Setup B.
-- **Setup B — Opus 4.6 Stability:** Opus 4.6 max driver on Max, 1M context included at standard rates (see above). Fable 5 advisor, Max-bundled. GPT-5.5 xhigh reviewer, separate.
+- **Setup A — Sonnet 5 + Fable advisor:** Sonnet 5's native 1M context — interactive session, Max-bundled, no `[1m]` suffix needed. Fable 5 advisor via `advisorModel: "fable"` — also Max-bundled. GPT-5.6 Sol xhigh reviewer on ChatGPT subscription. Roughly 5x less Max quota consumed per turn than Setup B.
+- **Setup B — Opus 4.6 Stability:** Opus 4.6 max driver on Max, 1M context included at standard rates (see above). Fable 5 advisor, Max-bundled. GPT-5.6 Sol xhigh reviewer, separate.
 - **Setup C — OpusPlan Hybrid:** **fully Max-bundled.** `opusplan` uses Opus (plan mode) + Sonnet (execute mode), both at their native context windows — no credit drain.
   - **⚠️ Avoid `sonnet[1m]` as a manual pin outside Setup A/C:** if your provider or gateway doesn't resolve Sonnet 5 to its native 1M automatically, forcing a `[1m]`-suffixed pin on an older Sonnet can draw from your usage credits pool ($3/$15 per Mtok) instead of your Max subscription. The `/model` picker shows this explicitly — watch for "Draws from usage credits."
-- **Reviewer (GPT-5.5 xhigh) in all three triads:** billed against your OpenAI account, completely separate from Anthropic.
+- **Reviewer (GPT-5.6 Sol xhigh) in all three triads:** billed against your OpenAI account, completely separate from Anthropic.
 - **CI loops that use `claude -p` post-June-15:** these now bill against the separate Anthropic credit pool, not your Max subscription. The wizard's CI shepherd loops (E2E scoring, weekly-update jobs) are local-only on the maintainer's machine and stay on Max; consumer-repo CI integrations may need to budget the new credit pool.
 
 ### Bottom line

diff --git a/CLAUDE.md b/CLAUDE.md
@@ -72,7 +72,7 @@ Key concepts:
 - PRs require 1 approving review and passing CI (`validate`)
 - E2E signal is advisory-only now, via `tests/e2e/local-shepherd.sh` run locally on maintainer's Max subscription (ROADMAP #212 Option 1)
 - Admin enforcement is on — no bypassing, even for repo owners
-- **Fable is the primary reviewer/advisor** (via `advisor()` or Fable subagent fallback when advisor is down). Codex (GPT-5.5) xhigh is the **cross-model safety check** — default to running before committing and pushing. Skip only with logged justification (e.g., single-line typo fix). Incident 2026-06-09: 4 PRs shipped without cross-model check, all had issues
+- **Fable is the primary reviewer/advisor** (via `advisor()` or Fable subagent fallback when advisor is down). Codex (GPT-5.6 Sol) xhigh is the **cross-model safety check** — default to running before committing and pushing. Skip only with logged justification (e.g., single-line typo fix). Incident 2026-06-09: 4 PRs shipped without cross-model check, all had issues
 
 ## Special Notes
 

diff --git a/CLAUDE_CODE_SDLC_WIZARD.md b/CLAUDE_CODE_SDLC_WIZARD.md
@@ -257,7 +257,7 @@ Claude Code's **effort level** controls how much thinking the model does before
 | Opus 4.8 (escalation only) | `xhigh` | `max` triggers excessive reasoning on 4.7/4.8 — documented 40-60x cache-token jump vs `high` (see "Opus 4.6 Stability" tier below) |
 | Fable 5 (advisor) | `high` (its own designed default) | Adaptive thinking always on; `xhigh`/`max` mainly move the thinking-token budget with small marginal gain for review work |
 | Opus 4.6 (Stability lane) | `max` | The one model where `max` doesn't overthink — no `xhigh` support at all (only low/medium/high/max) |
-| OpenAI/Codex (cross-model reviewer) | `xhigh` (their highest) | Always — lower reasoning misses subtle bugs the reviewer exists to catch |
+| OpenAI/Codex (cross-model reviewer) | `xhigh` default, escalate to `max`/Pro for unusually risky PRs | Lower reasoning misses subtle bugs the reviewer exists to catch; see `AI_SETUP_LANES.md`'s Final Review Policy for when to escalate |
 
 **Strict effort behavior (Opus 4.7+, carried forward in 4.8):**
 - **`xhigh` was introduced in 4.7** — sits between `high` and `max`, designed for coding and agentic work (30+ minute tasks with token budgets in the millions)
@@ -1087,7 +1087,7 @@ This is **Setup C (OpusPlan Hybrid/Saver)** in `AI_SETUP_LANES.md`. CC's native
 |-------|--------------------|--------------------|----------------------|
 | Planner | Opus 4.8 `xhigh` (Plan Mode) | Sonnet 5 `high`→`xhigh` | Opus 4.6 `max` |
 | Driver | Sonnet 5 `high` (execute mode) | Sonnet 5 `high`→`xhigh` | Opus 4.6 `max` |
-| Reviewer | GPT-5.5 xhigh | GPT-5.5 xhigh | GPT-5.5 xhigh |
+| Reviewer | GPT-5.6 Sol xhigh | GPT-5.6 Sol xhigh | GPT-5.6 Sol xhigh |
 
 **How to opt in:**
 ```json
@@ -1110,7 +1110,7 @@ Set effort per-session with `/effort` (planner `xhigh`, driver `high`) rather th
 **Prove-It Gate (#233 acceptance criterion):** mixed-mode ships only if pair-tested on 3+ simple repos shows Sonnet-coder + Opus-reviewer produces ≥ same SDLC scores as full-Opus baseline. The first version of the heuristic ships v1.38.0; pair-test results land in CHANGELOG before recommending mixed-mode as the default for any tier.
 
 **Tradeoffs (be honest):**
-- The Sonnet driver will drop some fine-grained self-review moves compared to an Opus-coder run — it's fast, less deliberate. The Opus planner and GPT-5.5 reviewer catch them, but expect more "fix in round 2" cycles.
+- The Sonnet driver will drop some fine-grained self-review moves compared to an Opus-coder run — it's fast, less deliberate. The Opus planner and GPT-5.6 Sol reviewer catch them, but expect more "fix in round 2" cycles.
 - Mixed-mode disables auto-mode (same as any pinned model). The pin is per-session — to switch back, remove the `model` line.
 
 ### Latest tier — Opus 4.8 (escalation model, #395)
@@ -3852,12 +3852,12 @@ Use an independent AI model from a different company as a code reviewer. The aut
 
 **Why this works:** Two AI systems from different companies (e.g., Claude writes, GPT reviews) provide adversarial diversity. They have fundamentally different training, different failure modes, and different strengths. What one misses, the other catches.
 
-**Use the best model at the deepest reasoning.** This is your quality gate — don't economize on it. Always use the latest, most capable model available (**GPT-5.5 if you have access**, otherwise GPT-5.4) at maximum reasoning effort (`xhigh` — this is non-negotiable, lower settings miss subtle errors). Cheaper/faster models miss things. The whole point is catching what the authoring model couldn't.
+**Use the best model at the deepest reasoning.** This is your quality gate — don't economize on it. Always use the latest, most capable model available (**GPT-5.6 Sol if you have access**, otherwise Terra) at `xhigh` reasoning effort (non-negotiable — lower settings miss subtle errors; preserve this baseline rather than downgrading it on a model bump). Cheaper/faster models miss things. The whole point is catching what the authoring model couldn't. For unusually risky PRs, escalating to `max` or Pro mode is defensible — see `AI_SETUP_LANES.md`'s Final Review Policy — but `xhigh` is the default; no published data shows the higher tiers catching meaningfully more real bugs on ordinary PR review.
 
 **Prerequisites:**
 - Codex CLI installed: `npm i -g @openai/codex`
 - OpenAI API key configured: `export OPENAI_API_KEY=...`
-- Codex CLI picks up your OpenAI account's best available model automatically. If you have GPT-5.5 access, `codex exec` uses it; otherwise it falls back to GPT-5.4. No config change needed on your side.
+- Codex CLI picks up your OpenAI account's best available model automatically. If you have GPT-5.6 Sol access, `codex exec` uses it; otherwise it falls back to Terra. No config change needed on your side.
 - This is a local workflow tool — not required for CI/CD
 
 **The Protocol:**

diff --git a/README.md b/README.md
@@ -123,7 +123,7 @@ export OPENAI_API_KEY=sk-...
 codex --version   # confirm ready
 ```
 
-That's it. Codex picks up your OpenAI account's best available model automatically — **if you have GPT-5.5, it uses GPT-5.5; otherwise GPT-5.4**. No model config needed.
+That's it. Codex picks up your OpenAI account's best available model automatically — **if you have GPT-5.6 Sol, it uses Sol; otherwise it falls back to Terra**. No model config needed.
 
 **How to use it:** after Claude's self-review passes, write a one-file mission brief and run:
 
@@ -175,17 +175,17 @@ Or pin in `.claude/settings.json`:
 { "model": "sonnet", "advisorModel": "fable", "effortLevel": "xhigh" }
 ```
 
-**Effort is model-aware, not blanket `max`.** Sonnet 5: `high` default, `/effort xhigh` for hard tasks. Opus 4.8: `xhigh` (its own `max` overthinks). Opus 4.6: `max` (its one `xhigh`-less sweet spot). Set per-session with `/effort`, not a shell-rc or settings env var — persisting effort that way silently overrides a later `/effort` change after you switch models (see `SDLC.md`'s Lessons Learned for a real incident this caused). OpenAI/Codex reviewer: `xhigh` (its highest).
+**Effort is model-aware, not blanket `max`.** Sonnet 5: `high` default, `/effort xhigh` for hard tasks. Opus 4.8: `xhigh` (its own `max` overthinks). Opus 4.6: `max` (its one `xhigh`-less sweet spot). Set per-session with `/effort`, not a shell-rc or settings env var — persisting effort that way silently overrides a later `/effort` change after you switch models (see `SDLC.md`'s Lessons Learned for a real incident this caused). OpenAI/Codex reviewer: `xhigh` default — escalate to `max`/Pro mode only for unusually risky PRs (see `AI_SETUP_LANES.md`'s Final Review Policy).
 
 ### Four Setup Lanes
 
 The wizard defines four AI coding setups in [`AI_SETUP_LANES.md`](AI_SETUP_LANES.md):
 
 | Lane | Advisor | Driver | Reviewer | Escalation |
 |------|---------|--------|----------|------------|
-| **A — Recommended** | Fable 5 (advisorModel) | Sonnet 5, `high`→`xhigh` | GPT-5.5 xhigh | Opus 4.8 xhigh or Fable review |
-| **B — Stability** | Fable 5 (advisorModel) | Opus 4.6 max | GPT-5.5 xhigh | None |
-| **C — Saver** | Fable 5 or Opus 4.8 (advisorModel) | Opus 4.8 plans, Sonnet 5 executes | GPT-5.5 xhigh | None |
+| **A — Recommended** | Fable 5 (advisorModel) | Sonnet 5, `high`→`xhigh` | GPT-5.6 Sol xhigh | Opus 4.8 xhigh or Fable review |
+| **B — Stability** | Fable 5 (advisorModel) | Opus 4.6 max | GPT-5.6 Sol xhigh | None |
+| **C — Saver** | Fable 5 or Opus 4.8 (advisorModel) | Opus 4.8 plans, Sonnet 5 executes | GPT-5.6 Sol xhigh | None |
 | **D — Lite** | None | Sonnet 5, `medium` | None | None |
 
 Setup D's whole point: **the discipline of knowing when NOT to use discipline.** When blast radius is low and you just need fast cheap hands, skip the SDLC overhead.