chore(models): move the opus alias to claude-opus-5

Claude Opus 5 released today — same $5/$25 sticker, 1M context; Opus 4.8
moves to legacy. The pricing table gains a dedicated claude-opus-5
fragment (the claude-opus-4 substring doesn't cover it, so without the
row the fleet's Opus usage would silently cost-track as $0 — exactly
what test_opus_is_priced now guards).

CHANGELOG.md

@@ -30,7 +30,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 
 ### Changed
 
-- **The `opus` alias upgrades from Claude Opus 4.6 to Claude Opus 4.8.** `MODEL_MAP["opus"]` now resolves to `claude-opus-4-8` — the newest Opus tier, at the same price as 4.6 — so every Opus-routed role (PR reviewer, Auditor, Board, CEO, Intake, Secretary) rides it at next spawn. Pricing already covered the id via the `claude-opus-4` fragment; a new `test_opus_is_priced` guard keeps the alias from ever pointing at an unpriced model.
+- **The `opus` alias upgrades from Claude Opus 4.6 to Claude Opus 5.** `MODEL_MAP["opus"]` now resolves to `claude-opus-5` — the newest Opus tier, released 2026-07-25 at the same $5/$25 price — so every Opus-routed role (PR reviewer, Auditor, Board, CEO, Intake, Secretary) rides it at next spawn. The pricing table gains a dedicated `claude-opus-5` fragment (the `claude-opus-4` substring doesn't cover it — without the row, Opus fleet usage would silently cost-track as $0), and a new `test_opus_is_priced` guard keeps the alias from ever pointing at an unpriced model.
 - **Panel dialogs converge on one shape: one disclosure primitive, DialogFooter everywhere, three dialog widths.** `collapsible-section` becomes the single sectioned-disclosure primitive — task dialogs' raw `Collapsible` and create-project's ad-hoc `showAdvanced` both move onto it; every hand-rolled dialog footer becomes `DialogFooter`; dialog widths collapse from ten ad-hoc classes to three named sizes, with deliberate outliers annotated. No behavioral change.
 
 ### Fixed

CLAUDE.md

@@ -79,7 +79,7 @@ pnpm test
 | RAG Engine | in-house (asyncpg + pgvector, hybrid retrieval) |
 | Cache/Queue | Redis |
 | Container Runtime | Docker + Docker Compose |
-| Cloud LLM | Claude API (claude-opus-4-8) + xAI Grok (official `grok` CLI, SuperGrok subscription) + OpenAI (official `codex` CLI, ChatGPT subscription) + Google Gemini (official `gemini` CLI, OAuth login) |
+| Cloud LLM | Claude API (claude-opus-5) + xAI Grok (official `grok` CLI, SuperGrok subscription) + OpenAI (official `codex` CLI, ChatGPT subscription) + Google Gemini (official `gemini` CLI, OAuth login) |
 | Local LLM | Ollama (glm-5.2:cloud for RAG/hybrid retrieval) |
 | Embeddings | qwen3-embedding:0.6b (1024 dim) |
 | Frontend | Next.js 16 + TypeScript + Tailwind + Radix UI (in `panel/`) |

docs/map/_complete_map.md

@@ -5134,7 +5134,7 @@ The Pydantic/dataclass domain surface of RoboCo — the typed contract the API,
 | `OrchestratorAgentState` | StrEnum | runtime.py:15 | offline/starting/active/waiting_short/waiting_long/idle/stopping |
 | `AgentInstance` | dataclass | runtime.py:70 | Running Claude Code container record + `usage_session_id` |
 | `SpawnGitContext` | dataclass | runtime.py:28 | Git context for spawn (project_slug, branch, `task_short_id` for worktree) |
-| `MODEL_MAP` | dict | runtime.py:106 | short-name→full Claude id (opus→claude-opus-4-8, sonnet→claude-sonnet-5, haiku→…) |
+| `MODEL_MAP` | dict | runtime.py:106 | short-name→full Claude id (opus→claude-opus-5, sonnet→claude-sonnet-5, haiku→…) |
 | `ROLE_MODEL_MAP` | dict | runtime.py:120 | per-role default tier: developer/qa/documenter/cell_pm/main_pm→sonnet, pr_reviewer/auditor/board/ceo/prompter/secretary→opus. **Haiku retired from every delivery-lifecycle role (#680, 2026-07-24)**: qa/documenter moved haiku→sonnet — haiku can't reliably emit the structured envelopes the lifecycle runs on (`pass_review`'s per-AC `criteria_verified`, `delegate`'s `covers_parent_criteria`, the findings ledger), so a haiku QA/PM claims, gets validation-rejected, idles, respawns, and loops without progress. A structured-verb capability floor (`_below_capability_floor`/`_floor_below_capability`, `roboco/services/llm.py`) additionally upgrades ANY below-floor Anthropic assignment to sonnet at resolution time — from a pin, a plain ROLE row, or a future map edit — so the floor can't be silently reintroduced. Non-Anthropic providers are untouched (an Anthropic-tier floor, not a provider policy). |
 | `ROLE_EFFORT_MAP` | dict | runtime.py | per-role `CLAUDE_CODE_EFFORT_LEVEL` override injected at spawn; **empty/inert by default** (opt-in per role after verifying the level moves usage) |
 | `MODEL_CATALOG` | tuple | llm_catalog.py:67 | Settings-dropdown source of truth; Anthropic entries derived from `MODEL_MAP` |

docs/map/models.md

@@ -83,7 +83,7 @@ The Pydantic/dataclass domain surface of RoboCo — the typed contract the API,
 | `OrchestratorAgentState` | StrEnum | runtime.py:15 | offline/starting/active/waiting_short/waiting_long/idle/stopping |
 | `AgentInstance` | dataclass | runtime.py:70 | Running Claude Code container record + `usage_session_id` |
 | `SpawnGitContext` | dataclass | runtime.py:28 | Git context for spawn (project_slug, branch, `task_short_id` for worktree) |
-| `MODEL_MAP` | dict | runtime.py:106 | short-name→full Claude id (opus→claude-opus-4-8, sonnet→claude-sonnet-5, haiku→…) |
+| `MODEL_MAP` | dict | runtime.py:106 | short-name→full Claude id (opus→claude-opus-5, sonnet→claude-sonnet-5, haiku→…) |
 | `ROLE_MODEL_MAP` | dict | runtime.py:120 | per-role default tier: developer/qa/documenter/cell_pm/main_pm→sonnet, pr_reviewer/auditor/board/ceo/prompter/secretary→opus. **Haiku retired from every delivery-lifecycle role (#680, 2026-07-24)**: qa/documenter moved haiku→sonnet — haiku can't reliably emit the structured envelopes the lifecycle runs on (`pass_review`'s per-AC `criteria_verified`, `delegate`'s `covers_parent_criteria`, the findings ledger), so a haiku QA/PM claims, gets validation-rejected, idles, respawns, and loops without progress. A structured-verb capability floor (`_below_capability_floor`/`_floor_below_capability`, `roboco/services/llm.py`) additionally upgrades ANY below-floor Anthropic assignment to sonnet at resolution time — from a pin, a plain ROLE row, or a future map edit — so the floor can't be silently reintroduced. Non-Anthropic providers are untouched (an Anthropic-tier floor, not a provider policy). |
 | `ROLE_EFFORT_MAP` | dict | runtime.py | per-role `CLAUDE_CODE_EFFORT_LEVEL` override injected at spawn; **empty/inert by default** (opt-in per role after verifying the level moves usage) |
 | `MODEL_CATALOG` | tuple | llm_catalog.py:67 | Settings-dropdown source of truth; Anthropic entries derived from `MODEL_MAP` |

panel/src/components/settings/__tests__/ai-routing-card.test.tsx

@@ -33,9 +33,9 @@ const {
 } = vi.hoisted(() => ({
   catalog: vi.fn(async () => [
     {
-      model_name: "claude-opus-4-8",
+      model_name: "claude-opus-5",
       provider_type: "anthropic",
-      display_name: "Claude Opus 4.8",
+      display_name: "Claude Opus 5",
     },
     {
       model_name: "grok-build-0.1",
@@ -719,14 +719,14 @@ describe("AIRoutingCard", () => {
 
       const devLow = screen.getByTestId("complexity-select-developer-low");
       fireEvent.click(
-        await within(devLow).findByRole("option", { name: "Claude Opus 4.8" }),
+        await within(devLow).findByRole("option", { name: "Claude Opus 5" }),
       );
 
       await waitFor(() =>
         expect(setComplexityOverride).toHaveBeenCalledWith({
           role: "developer",
           complexity: "low",
-          model_name: "claude-opus-4-8",
+          model_name: "claude-opus-5",
         }),
       );
     });
@@ -764,7 +764,7 @@ describe("AIRoutingCard", () => {
 
       const devLow = screen.getByTestId("complexity-select-developer-low");
       fireEvent.click(
-        await within(devLow).findByRole("option", { name: "Claude Opus 4.8" }),
+        await within(devLow).findByRole("option", { name: "Claude Opus 5" }),
       );
 
       await waitFor(() => expect(setComplexityOverride).toHaveBeenCalled());

panel/src/components/tg/tg-metrics-demo.ts

@@ -105,7 +105,7 @@ export const DEMO_TEAM_USAGE: TeamUsageRow[] = (
 
 export const DEMO_MODEL_USAGE: ModelUsageSlice[] = (
   [
-    ["claude-opus-4-8", 42.1],
+    ["claude-opus-5", 42.1],
     ["glm-5.2:cloud", 16.8],
     ["grok-build", 7.64],
   ] as Array<[string, number]>

roboco/billing/pricing.py

@@ -57,11 +57,14 @@ _ANTHROPIC_FRAGMENTS = ("claude", "opus", "sonnet", "haiku")
 # Cache read is charged at ~10 % of the input price.
 # Cache write is charged at ~25 % of the input price.
 #
-# Match on *substring* of model name so "claude-opus-4-8" and "opus" both
+# Match on *substring* of model name so "claude-opus-5" and "opus" both
 # resolve to the same tier.
 # ---------------------------------------------------------------------------
 _PRICING: list[tuple[str, float, float, float, float]] = [
     # (fragment, input/1M, output/1M, cache_read/1M, cache_write/1M)
+    # Opus 5 — same sticker as the Opus 4 family; needs its own fragment
+    # because "claude-opus-4" doesn't substring-match it.
+    ("claude-opus-5", 5.00, 25.00, 0.50, 6.25),
     # Opus 4 family
     ("claude-opus-4", 5.00, 25.00, 0.50, 6.25),
     # Sonnet 4 / 3.7 / 3.5 family

roboco/models/llm_catalog.py

@@ -7,7 +7,7 @@ the router maps back to the correct pre-seeded provider row.
 
 **Anthropic entries derive from `runtime.MODEL_MAP`** — that's the single
 source of truth for which Claude versions are supported. Bumping a model
-version there (e.g. `claude-opus-4-8` → a newer id) updates the
+version there (e.g. `claude-opus-5` → a newer id) updates the
 catalog automatically without a second edit here.
 
 Ollama Cloud entries are hand-maintained because Ollama's cloud tags
@@ -46,7 +46,7 @@ def _build_anthropic_entries() -> tuple[CatalogEntry, ...]:
 
     Keeps the UI honest: a user picking "Claude Opus" sees exactly which
     underlying Claude Code model id will be used at spawn
-    (e.g. "Claude Opus · claude-opus-4-8"), so version bumps in
+    (e.g. "Claude Opus · claude-opus-5"), so version bumps in
     `runtime.MODEL_MAP` are immediately visible.
     """
     entries: list[CatalogEntry] = []

roboco/models/runtime.py

@@ -110,7 +110,7 @@ class WaitingRecord:
 
 # Model mapping for cost optimization
 MODEL_MAP: dict[str, str] = {
-    "opus": "claude-opus-4-8",
+    "opus": "claude-opus-5",
     "sonnet": "claude-sonnet-5",
     "haiku": "claude-haiku-4-5-20251001",
 }

tests/unit/billing/test_pricing.py

@@ -565,6 +565,7 @@ class TestProviderAwareness:
         for name in (
             "claude-opus-4-6",
             "claude-opus-4-8",
+            "claude-opus-5",
             "claude-fable-5",
             "opus",
             "sonnet",

tests/unit/models/test_misc_models.py

@@ -106,7 +106,7 @@ def test_build_anthropic_entries_skips_missing_models() -> None:
     # Patch MODEL_MAP to be missing one entry → loop hits the `continue` branch.
     with patch(
         "roboco.models.llm_catalog.MODEL_MAP",
-        {"opus": "claude-opus-4-8"},  # sonnet+haiku missing
+        {"opus": "claude-opus-5"},  # sonnet+haiku missing
     ):
         entries = _build_anthropic_entries()
     # Only one entry survives — opus.

tests/unit/models/test_model_map.py

@@ -12,10 +12,9 @@ def test_sonnet_alias_resolves_to_sonnet_5() -> None:
     assert MODEL_MAP["sonnet"] == "claude-sonnet-5"
 
 
-def test_opus_alias_resolves_to_opus_4_8() -> None:
+def test_opus_alias_resolves_to_opus_5() -> None:
-    # CEO upgrade to the newest Opus tier ("Opus 5" does not exist — the
+    # CEO upgrade to the newest Opus tier (Opus 5, released 2026-07-25).
-    # Claude 5 family is Fable/Mythos/Sonnet; Opus tops out at 4.8).
+    assert MODEL_MAP["opus"] == "claude-opus-5"
-    assert MODEL_MAP["opus"] == "claude-opus-4-8"
 
 
 def test_sonnet_5_is_priced() -> None:

tests/unit/runtime/test_intake_spawn.py

@@ -56,7 +56,7 @@ def _spec(**overrides: Any) -> _IntakeRunSpec:
         },
         "session_id": "sess-abc",
         "cwd": "/data/workspaces/roboco/board/intake-1",
-        "cli_model": "claude-opus-4-8",
+        "cli_model": "claude-opus-5",
         "api_url": "http://roboco-orchestrator:8000",
         "provider_base_url": None,
         "provider_auth_token": None,
@@ -108,7 +108,7 @@ class TestBuildIntakeRunCmd:
         assert "ROBOCO_WORKSPACE=/data/workspaces/roboco/board/intake-1" in cmd
         assert "ROBOCO_API_URL=http://roboco-orchestrator:8000" in cmd
         assert "ROBOCO_AGENT_ID=intake-1" in cmd
-        assert "CLAUDE_CODE_SUBAGENT_MODEL=claude-opus-4-8" in cmd
+        assert "CLAUDE_CODE_SUBAGENT_MODEL=claude-opus-5" in cmd
 
     def test_mounts_prompt_and_workspaces(self) -> None:
         cmd = AgentOrchestrator._build_intake_run_cmd(_spec())

tests/unit/runtime/test_orchestrator_manifest.py

@@ -99,7 +99,7 @@ class TestBuildManifestForAgent:
             mock_settings.manifest_host_dir = str(tmp_path)
             mock_settings.workspaces_root = str(tmp_path / "workspaces")
 
-            result = _build_manifest_for_agent("be-dev-1", "claude-opus-4-8")
+            result = _build_manifest_for_agent("be-dev-1", "claude-opus-5")
 
         assert result is not None
         data = json.loads(result.read_text())

tests/unit/runtime/test_spawn_manifest.py

@@ -87,7 +87,7 @@ class TestBuildForRole:
                 role=role,
                 team="board",
                 workspace_path=Path("/tmp/x"),
-                agent_model="claude-opus-4-8",
+                agent_model="claude-opus-5",
             )
         )
         assert "read_messages" in m.do_tools

tests/unit/services/test_audit_agent_id.py

@@ -182,7 +182,7 @@ async def test_log_agent_event_resolves_slug_to_uuid(
         event_type="agent.spawned",
         agent_slug=slug,
         task_id=task_id,
-        details={"container_id": "abc123def456", "model": "claude-opus-4-8"},
+        details={"container_id": "abc123def456", "model": "claude-opus-5"},
     )
 
     # Read back the row.

tests/unit/services/test_usage.py

@@ -945,7 +945,7 @@ class TestGetRecentSessions:
         row = MagicMock()
         row.id = sid
         row.agent_slug = "product-owner"
-        row.model = "claude-opus-4-8"
+        row.model = "claude-opus-5"
         row.started_at = datetime.datetime(2026, 6, 11, 20, 41, tzinfo=datetime.UTC)
         row.ended_at = datetime.datetime(2026, 6, 11, 20, 42, tzinfo=datetime.UTC)
         row.tokens_input = exp_in
@@ -966,7 +966,7 @@ class TestGetRecentSessions:
         s = out[0]
         assert s["id"] == str(sid)
         assert s["agent_slug"] == "product-owner"
-        assert s["model"] == "claude-opus-4-8"
+        assert s["model"] == "claude-opus-5"
         assert s["tokens_input"] == exp_in
         assert s["tokens_output"] == exp_out
         assert s["tokens_cache"] == exp_cr + exp_cw