sortie-dogs 0.13.11 → 0.13.13

package/README.md

@@ -4,729 +4,236 @@
   <img src="docs/assets/sortie-dogs-logo.png" alt="Sortie-dogs logo" width="640">
 </p>
 
-**A goal-preserving, adaptive execution harness for OpenCode that optimizes cost,
-time, and proof without taking your setup over.**
-
-Use OpenCode normally. Invoke Sortie only when you want scoped investigation,
-implementation, validation, review, and model routing.
-
-- **Goal invariance**: accepted outcomes and proof requirements survive delegation,
-  continuation, remediation, and restart.
-- **Adaptive execution**: small work stays small; additional agents and stronger
-  models are used only when task shape or risk justifies them.
-- **Clear Worker instructions**: give Luna a concise goal, explicit constraints
-  and completion criteria; keep orchestration bookkeeping in the harness.
-  See the [instruction design principles](docs/worker-instruction-design.md).
-- **Coexistence**: Sortie activates only when selected and preserves normal
-  OpenCode agents, settings, and user-owned files.
-- **Cost, time, and proof**: the objective is a verified result at the lowest
-  practical cost and wall time, not the largest agent count.
+**An adaptive execution harness for OpenCode and Codex: preserve the goal,
+implement, validate, review, and deliver without taking over your setup.**
+
+Use your host normally. Invoke Sortie when a task needs coordinated execution and an
+evidence-backed result, not more agents for their own sake.
 
 [![GitHub Release](https://img.shields.io/github/v/release/zufall-upon/Sortie-dogs)](https://github.com/zufall-upon/Sortie-dogs/releases/latest)
 [![npm](https://img.shields.io/npm/v/sortie-dogs?label=npm)](https://www.npmjs.com/package/sortie-dogs)
 [![Tests](https://github.com/zufall-upon/Sortie-dogs/actions/workflows/test.yml/badge.svg)](https://github.com/zufall-upon/Sortie-dogs/actions/workflows/test.yml)
-[![OpenCode Plugin](https://img.shields.io/badge/OpenCode-Plugin-5C5CFF)](https://opencode.ai/)
+[![OpenCode](https://img.shields.io/badge/OpenCode-Plugin-5C5CFF)](https://opencode.ai/)
+[![Codex](https://img.shields.io/badge/Codex-Host_adapter-222222)](docs/codex.md)
 [![TypeScript](https://img.shields.io/badge/TypeScript-5.9-3178C6?logo=typescript&logoColor=white)](https://www.typescriptlang.org/)
 [![Node.js](https://img.shields.io/node/v/sortie-dogs)](https://www.npmjs.com/package/sortie-dogs)
 [![MIT License](https://img.shields.io/npm/l/sortie-dogs)](LICENSE)
 
-![Sortie-dogs coordinating a bounded implementation workflow](https://raw.githubusercontent.com/zufall-upon/Sortie-dogs/main/docs/assets/sortie-workflow.gif)
-
-Guides: [日本語](docs/guide-ja.md) · [简体中文](docs/guide-zh-CN.md) ·
-[Testing](docs/testing.md) · [CLI testing](docs/cli-testing.md)
-
-**Current release: [v0.13.10](https://github.com/zufall-upon/Sortie-dogs/releases/tag/v0.13.10)**
-([release notes](docs/release-v0.13.10.md)). The default Mission runtime retains the `v010`
-profile, command and configuration names for compatibility; these names do not mean v0.10 is installed.
-The current asset marker is `0.13.10-codex-windows-v1`.
-
-## SWE-bench Lite: 170/300 (56.67%)
-
-The fixed **Sortie-dogs v0.12.24** harness resolved **170 of 300 SWE-bench Lite test issues** in one pass@1 campaign, with 9 empty patches and no official evaluation errors. Every instance has a frozen prediction and an inference-time trajectory. The task Workers ran `openai/gpt-6-luna-fast#max`; operator, coordinator and review roles ran `openai/gpt-6-sol#xhigh`. This is a system result, **not** a Luna-only model comparison or a Verified/full SWE-bench score.
+**Current release: [v0.13.13](https://github.com/zufall-upon/Sortie-dogs/releases/tag/v0.13.13)**
+· [Release notes](docs/release-v0.13.13.md)
 
-[Technical report and per-repository results](docs/benchmarks/swebench-lite-v01224-test300-2026-09-29.md) · [Public predictions, logs and trajectories](https://github.com/zufall-upon/sortie-dogs-swebench-lite-20260929)
+[Quick start](#quick-start) · [Workflow](#mission-workflow) · [Models](#default-models) ·
+[Measured results](#measured-results) · [Configuration](#configuration) · [Documentation](#documentation)
 
-The single official 300-instance report and frozen predictions are hash-bound in the report. Confirmed inference expense was **$162.99**; a separate **$34.60** of usage has unknown pricing and is held against the campaign cap, **not** counted as known expense. Leaderboard registration and maintainer acceptance are separate from this official local evaluation.
+> **Beta:** the v0.13.x package is still stabilizing before 1.0. Both host integrations are
+> implemented; Codex Missions have completed implementation, validation, correction and acceptance.
+> [Host-specific execution boundaries](docs/codex.md#native-permissions-and-approval) remain explicit.
 
-Historical scores below belong to their fixed candidates, not v0.13.10. SWE-bench is a separate,
-optional measurement rather than a mandatory release gate.
+![Sortie-dogs coordinating a bounded implementation workflow](https://raw.githubusercontent.com/zufall-upon/Sortie-dogs/main/docs/assets/sortie-workflow.gif)
 
-> **Beta:** v0.13.x is still stabilizing. Runtime behavior,
-> configuration, and generated assets may still change before 1.0.
+## Why Sortie
+
+- **Autonomy:** retain the original request across delegation, correction, compaction and restart.
+  Finish in-request work through the existing host authority instead of routine approval round trips.
+- **Efficiency:** small work stays small. Use a direct Worker for a known single unit; add
+  Coordinator, Scout or Advisor only when discovery or task shape calls for them. Reuse valid
+  unchanged evidence instead of repeating successful work.
+- **Visibility:** show real commands, exits, timing, model routes, review disposition and final
+  acceptance. Unknown execution or cost stays unknown, not an invented success or zero.
+- **Quality:** meaningful formal checks and risk-based review establish completion, not model prose.
+  Worker instructions carry a concise goal, explicit constraints and completion criteria;
+  the harness owns bookkeeping. [Instruction design](docs/worker-instruction-design.md).
+- **Coexistence:** explicit invocation, existing authentication and host settings, one shared
+  Mission lifecycle. Normal OpenCode/Codex use and user-owned files remain yours.
 
 ## Quick start
 
-### Codex host adapter (experimental)
+Choose **one host route**. Both require Node.js **22.6+** and npm; project-local installation is
+recommended. OpenCode is not required for Codex, and Codex is not required for OpenCode.
 
-Choose this route for Codex; OpenCode is not required and **do not run `sortie-dogs init` for it**.
-Use Node.js 22.6 or newer and an existing Codex CLI with ChatGPT authentication. Ubuntu validation
-used Node.js 22.22.1 and Codex 0.160.1. Sortie neither installs Codex nor starts a login flow, copies
-credentials, or creates a second host configuration. Mission execution refuses non-ChatGPT auth.
+### Codex
 
-The Codex Mission adapter and explicit skill are included in v0.13.10. Install the release in the
-target project and inspect the entrypoint before running a prompt:
+Use an existing Codex CLI signed in with **ChatGPT**. On Windows, PowerShell **7** must already be
+available; POSIX uses bash. Sortie does not install Codex, start login or use a metered API fallback.
 
 ```sh
-npm install --save-dev sortie-dogs@0.13.10
+npm install --save-dev sortie-dogs@0.13.13
 npx --no-install sortie-dogs codex init .
-npx --no-install sortie-dogs codex mission --help
 ```
 
-For an unpublished development candidate, a maintainer with checkout dependencies already present
-can produce a local tarball with `npm pack` (build included) and install that tarball instead.
-This does not publish it. Use the same installed version for the CLI and SDK below.
-`codex init` installs only `.agents/skills/sortie-dogs`; it does not create or edit `.opencode`.
-It refuses to overwrite a skill whose ownership cannot be established.
-
-After restarting Codex or opening the project in a new chat, invoke the installed skill explicitly:
+Restart Codex or open the project in a new chat, then invoke the installed skill explicitly:
 
 ```text
 $sortie-dogs Implement and verify the requested change
 ```
 
-`$sortie-dogs` is the current direct Skills syntax. Codex does not expose arbitrary custom commands
-as exact `/sortie-dogs` slash commands; `/skills` opens the skill picker. The skill is intentionally
-explicit-only so child Codex Mission turns cannot invoke another Sortie Mission recursively.
-On POSIX and Windows it selects the natural-language, multi-role Mission route below.
-Windows uses existing PowerShell 7; it does not route natural-language Missions through the
-single-task `codex run` manifest workflow. That workflow remains available when explicitly selected
-and retains its post-execution observation boundary, not a pre-execution guard.
-
-Sortie-dogs also exports a host adapter for the official Codex app-server stdio protocol. This is
-separate from selecting an OpenAI model through OpenCode: OpenCode model routing still uses the
-OpenCode host, while `CodexAppServerHost` starts and observes native Codex threads and turns.
-
-```ts
-import { CodexAppServerHost, createCodexAppServerTransport } from "sortie-dogs";
-
-const transport = createCodexAppServerTransport({ executable: "codex" });
-const host = new CodexAppServerHost(transport);
-const threadID = await host.startThread({ cwd: process.cwd(), ephemeral: true });
-const result = await host.runTurn(threadID, "Implement the admitted unit.", {
-  cwd: process.cwd(),
-  approvalPolicy: "unlessTrusted",
-  sandboxPolicy: {
-    type: "workspaceWrite",
-    writableRoots: [process.cwd()],
-    networkAccess: false,
-  },
-});
-await host.close();
-```
-
-The adapter reuses Sortie's host-neutral core instead of copying its goal, acceptance, evidence, or
-ledger implementation. It exposes completed Codex items as authoritative observations, supports
-exact persisted-thread resume and turn interruption (ephemeral threads have no resumable rollout), and declines command/file approvals unless the caller
-provides an approval handler. Permission-subset requests receive an empty grant unless the caller
-connects `permissionsApproval`; approved responses are limited to the native request. Unsupported
-server-initiated requests fail closed.
-
-For the existing Operator → Coordinator → Worker → Reviewer Mission workflow, use:
+Or run the same natural-language Mission from the terminal:
 
 ```sh
 npx --no-install sortie-dogs codex mission --prompt "Implement and review the requested change"
 ```
 
-The equivalent public SDK entrypoint is:
-
-```ts
-import { CodexMissionSession } from "sortie-dogs";
-
-const mission = await CodexMissionSession.create({
-  projectRoot: process.cwd(),
-  // Leave model/effort unset to preserve packaged role defaults.
-  // model: "gpt-6.1-sol", effort: "medium", // Explicit override for every role.
-  // trustedPowerShellExecutable: "C:/Program Files/PowerShell/7/pwsh.exe", // Windows, optional.
-  // resumeThreadID: "saved-root-thread-id", // Continue the same Mission when needed.
-});
-try {
-  const result = await mission.run("Implement and review the requested change");
-  console.log(JSON.stringify(result));
-  process.exitCode = result.accepted ? 0 : 1;
-} finally {
-  await mission.close();
-}
-```
+`codex init` installs `.agents/skills/sortie-dogs` only; it does not edit `.opencode`.
+**Do not run `sortie-dogs init` for this route.** `$sortie-dogs` is Skills syntax, not an exact
+`/sortie-dogs` slash command. Leave model overrides unset to retain the role defaults.
+
+Existing native permissions remain authoritative. An SDK application can connect its already
+authorized parent-host executor; the CLI does not supply one or an interactive native approval
+bridge. [Codex guide: SDK, permissions, progress, resume and manifest route](docs/codex.md).
+
+### OpenCode
 
-This route uses `CodexMissionSession` and the existing Mission tools, correction grants,
-validation evidence, and final Operator acceptance. It runs saved native Codex threads with existing
-ChatGPT authentication; it does not import OpenCode settings or introduce a second Mission ledger.
-Commands run through `/bin/bash` on POSIX or existing `pwsh.exe -NoProfile -NonInteractive -Command`
-on Windows, using the native app-server's configured permissions. The compatibility tool remains
-named `bash`, but its Windows input must be PowerShell syntax. Windows callers can pin an absolute
-PowerShell 7 path with SDK `trustedPowerShellExecutable` or CLI `--trusted-pwsh`.
-Packaged roles retain Sol 6.1/xhigh for Operator, Coordinator and Reviewer. The shared Luna-fast/max
-Worker alias maps only in Codex to native `gpt-6-luna`/max plus `serviceTier: "priority"` (Fast).
-OpenCode routing is unchanged. Fast uses subscription limits faster than Standard; no metered API
-fallback is introduced. CLI progress exposes each role's native model, effort and separate service tier.
-Sortie does not
-replace them with a fixed repository-only or network-disabled policy, change host configuration, or select
-full access. Thread turns retain native thread permissions; standalone `command/exec` uses the server's
-configured policy, not a thread's temporary grants. A parent application's in-memory approval is not
-automatically transferred to a separately launched app-server.
-Windows recovery can reclaim a killed adapter only when its recorded Windows PID is absent.
-A live/reused PID, access denial, legacy owner without platform identity, or unresolved native
-execution remains unknown; no command is replayed. Sandbox startup errors must be resolved at the
-native host, not by silently widening Sortie's permissions or stopping unrelated Codex processes.
-SDK hosts can forward native command/file approval requests through `approval` and permission-subset
-requests through `permissionsApproval`. The native host remains responsible for deciding and enforcing
-the grant. The CLI reports approval requests but has no interactive approval bridge; without a connected
-host callback, no grant is returned. `command/exec` has no thread-scoped approval API, so inheriting the
-configured policy does not add an escalation path for these standalone commands.
-Use SDK `permissions` or CLI `--permissions <native-profile>` only to explicitly select an existing
-native profile for thread start/resume and standalone commands. Omission retains native defaults,
-which may be read-only. Invalid or disallowed profiles fail through the native server without fallback;
-no configuration file is changed. Low-level `CodexAppServerHost` callers must opt into
-`experimentalApi: true` for profile APIs; Mission sessions already negotiate that capability. Progress reports the effective native profile, sandbox and approval
-routing. This selection does not configure a delegated parent-host executor's own permissions.
-See [Ubuntu acceptance evidence and remaining work](docs/codex-mission-acceptance-20261006.md).
-
-For a parent application with an existing approval-aware executor, pass the optional SDK
-`executeCommand` callback. It receives the exact post-hook argv, cwd, timeout, thread/turn/call identity,
-and an AbortSignal for bash/read/write operations. The parent owns approval and execution; Sortie does
-not infer a grant from prompt text or create another permission store. Return `completed` only after
-execution ends, with the real integer exitCode, stdout and stderr. Approval alone is not a result.
-`denied` and `not-started` explicitly mean nothing ran and create no validation exit. `interrupted`,
-`unknown`, exceptions and invalid results retain unknown execution and stop resends. The host owns
-command timeout enforcement; Sortie does not add a separate deadline to its approval interaction.
-Closing the adapter signals cancellation; late callback results cannot establish validation, and the
-host must reconcile any still-running process. Omission keeps native `command/exec`; a connected
-executor never falls back to native execution after rejection or failure. Existing Mission before/after
-hooks, command identity and validation acceptance remain shared. CLI progress identifies the selected
-executor; the CLI itself does not attach a parent executor.
-File writes use a short argv referencing a temporary UTF-8 payload, so large files and literal quotes,
-newlines or NUL characters do not overflow Windows command-line limits. Staging does not write the
-project file; the same configured executor performs that write, and the payload is removed afterward.
-Status observations (without `confirmed_conditions`) can return during an active child Task. Condition
-registration and other modifying tools remain serialized; status does not imply Task completion.
-The bash tool accepts `workdir` (absolute or relative to the project root). The shared hooks and both
-executors use that directory; progress and receipts include the effective cwd. A successful command in
-another directory does not satisfy a validation declared for the project root.
-
-Packaged role models and reasoning levels apply by default. `--model` and `--effort` explicitly
-override all roles; SDK callers can use `roleModels` for individual roles. Unsupported native model
-names fail rather than silently selecting another model. Availability depends on the signed-in account.
-
-Use `--resume <root-thread-id>` to continue a saved Mission root. Without it, a single unfinished Codex
-Mission in this repository is selected automatically; multiple unfinished roots require an explicit
-selection. Resume restores native messages and tool evidence, including child sessions. Completed native
-receipts feed the existing reservation reconciliation without replaying tool execution. A new Task stopped
-before child binding can be recovered when the exact terminal parent thread/turn/call/input matches and
-the prior adapter is closed or its Linux process identity proves it is gone. The host retains that proof in
-the existing Mission and reconciles the reservation before another model prompt; repeated recovery does
-not settle it again. Concurrent recovery claims are serialized.
-
-A newly created ordinary Worker also saves its exact dispatch binding before the child prompt is sent.
-If its parent Task result is lost, recovery can match that binding to the child's single, fully loaded,
-completed native turn. The initial recovery path supports leaf Workers only: later turns, nested Tasks,
-untracked native children, missing bindings and ambiguous identities remain unknown. Recovery restores
-execution completion for the existing settlement path; missing validation evidence is a process defect,
-not an invented PASS. It does not rerun the implementation. Continue with only the work or validation
-still needed for normal Operator acceptance.
-
-SIGINT/SIGTERM closes the adapter and exits 130/143. Bound children without an exact completed dispatch,
-and commands sent without a saved terminal receipt, remain unknown and are not resent. App-server exit
-does not prove an external command has stopped. Linux process-death checks do not prove external writer
-quiescence either; other platforms require a clean adapter close for automatic prelaunch recovery.
-Inspect the original executor before recovery; do not delete Mission state to force a retry.
-Per-session usage is native thread cumulative usage (including earlier turns), not a Mission total;
-monetary cost is unavailable and reported as `null`.
-The existing Mission report uses observed per-turn token deltas and execution times. Cache and reasoning
-subtotals are counted once. Its pre-terminal snapshot excludes final-answer generation, while final JSON
-retains native thread totals. Terminal accounting observations survive cold reload in the existing Mission;
-missing baselines remain unavailable. Native turn aggregates do not establish model-request counts, API
-costs or remaining subscription allowance, so these values are not inferred.
-CLI exit 0 means Operator acceptance; exit 1 means incomplete work or an execution error; exit 2
-means invalid arguments. A native turn completing alone is not acceptance.
-The CLI writes bounded progress records to stderr: tool start/completion, commands and exit codes,
-replan reasons, next actions, child identity, and public model commentary. Final JSON remains on stdout.
-
-For a manifest-bound mission, use the SDK `runCodexMission(...)` or the CLI:
+Use **OpenCode V2**. In the target project:
 
 ```sh
-sortie-dogs codex run --manifest operation-manifest.json --prompt "Implement the requested change"
+npm install --save-dev sortie-dogs@0.13.13
+npx --no-install sortie-dogs init .
 ```
 
-The operation manifest remains required. The mission runner reuses Sortie's existing goal ledger,
-protected evidence capture, settlement, and repository-wide scope leases. It does not emulate
-OpenCode plugin hooks, install Codex assets, switch models implicitly, or copy/import/modify OpenCode
-settings and sessions. Codex command and file events are post-execution observations, not a claimed
-pre-execution write guard. The stdio transport is the supported initial integration; experimental
-WebSocket transport is intentionally out of scope.
-
-The manifest-run CLI handles SIGINT/SIGTERM by requesting interruption and cleaning up its
-app-server (exit 130/143). SDK callers can pass an `AbortSignal` as `signal`. Before
-turn dispatch, cancellation settles the reservation and releases its lease. After
-dispatch, app-server exit or an interrupt acknowledgement does not prove that an
-externally hosted command stopped. Cancellation or transport loss therefore preserves
-the active goal/reservation as an unknown outcome and stops lease heartbeats without
-claiming completion. SIGKILL can leave the same unresolved state.
+Completely restart OpenCode, then run:
 
-Lease expiry alone does not authorize a retry: subsequent Codex missions refuse with
-`codex-mission-outcome-unknown:no-resend` and the ledger path, including across
-stable/v010 profiles. Inspect the prior worker and ledger before recovery; do not
-delete the ledger or automatically resend the prompt. This is a Codex admission
-check; it does not establish quiescence of external command executors.
+```text
+/sortie-v010 Implement and verify the requested change
+```
 
-### OpenCode plugin
+Selecting `dog-operator` starts the same workflow. `init` registers the V2 plugin, installs Mission
+assets and sets subagent depth to at least two while preserving unrelated settings. Existing
+`sortie-dogs/server` bridges are reused. Internal `dogs-coordinator` / `*-v010` roles are not entrypoints.
 
-Requirements: Node.js 22.6 or newer, npm, and OpenCode V2. These installation and `init` steps
-configure the OpenCode route; they are not prerequisites for the Codex route above.
+The `v010` command, profile and configuration names remain for compatibility; they do **not** mean
+v0.10 is installed. [OpenCode setup and configuration](docs/configuration.md) ·
+[日本語](docs/guide-ja.md) · [简体中文](docs/guide-zh-CN.md).
 
-Run these commands in the target project:
+## Mission workflow
 
-```sh
-npm install --save-dev sortie-dogs@latest
-npx sortie-dogs init .
+```text
+Known unit: User → Operator → Worker → Validation → Review / Correction → Acceptance
+Discovery:  User → Operator → Coordinator → Worker → Validation → Review / Correction → Acceptance
 ```
 
-`init` defaults to the `v010` Mission profile and registers the OpenCode V2 plugin in
-`.opencode/opencode.json(c)`, preserving unrelated settings. It sets subagent depth to at least two:
+1. **Operator** saves original requirements and owns user decisions and final acceptance.
+2. **Worker** investigates, implements, runs meaningful checks and performs requested Git delivery
+   in the same useful unit. A known unit/check can go directly to Worker without a planning handoff.
+3. **Coordinator**, when needed, investigates or decomposes work, reconciles in-request scope and
+   dispatches units. It can also implement and validate directly.
+4. **Review** is initially independent for high-risk changes; low-risk skips are explicit.
+   Findings can be corrected and validated by the same Reviewer, followed by recorded author
+   self-recheck. That is not a second independent PASS.
+5. **Acceptance** compares the original request with actual source, checks and review evidence.
+   Only the succeeded receipt authorizes DONE and the measured return report.
 
-```json
-{
-  "plugins": ["sortie-dogs"],
-  "experimental": { "subagent_depth": 2 }
-}
-```
+The default Mission profile is serial. Background responsiveness and status observations do not
+add parallel writers, replay unknown commands or imply completion. More agents are not the goal.
 
-If an existing local bridge already imports `sortie-dogs/server`, `init` reuses it instead of adding
-a duplicate package entry. A larger existing subagent depth is retained.
+For **run-once / result-collection** requests, a terminal failure can be the requested result:
+validate and report it without an unauthorized repair/rerun loop. Acceptance preserves the real
+operation exit and `execution-failed` outcome; it never turns a failed benchmark into a successful
+one. Required success checks, missing commands and active/unstarted operations still block completion.
 
-Completely restart OpenCode, then run:
+[Mission tools and validation policy](docs/configuration.md#mission-tools) ·
+[Operation efficiency](docs/mission-operation-efficiency.md) ·
+[Review policy](docs/quality-first-review.md)
 
-```text
-/sortie-v010 <task>
-```
+## Default models
 
-Selecting `dog-operator` directly starts the same workflow. `dog-operator` is the
-user-facing entry point for the Mission profile. `dogs-coordinator` and every `*-v010` role are
-internal children and must not be selected as task entry points.
-
-`init` installs runtime assets and merges the required OpenCode settings; the package entry or
-existing local bridge loads enforcement and model routing. OpenCode can reload watched configuration,
-but replacing an installed dependency may require a full restart. A new chat session alone does not
-prove the newly installed plugin is loaded.
-
-## v0.13.10 runtime updates
-
-PR #169 connects natural-language Windows Codex Missions to the same multi-role lifecycle using
-existing PowerShell 7, preserves command failure exits and native resume settings, and maps the
-Luna-fast Worker alias to native Luna/max with the separate priority tier. Directory reads list
-direct entries instead of requiring filename guesses. OpenCode model settings and native permission
-profiles remain unchanged. PR #168 provides one host-aware Anko runner, reuses unchanged setup,
-removes mandatory paid diagnostic probes, and preserves prior attempts, unknown usage and service
-evidence. Its recorded v0.13.9 Anko completion and PR #169's Windows Codex host-executor completion
-are historical fixed-candidate observations, not new v0.13.10 benchmark or native sandbox results.
-See [release notes](docs/release-v0.13.10.md).
-
-## v0.13.9 runtime updates (retained)
-
-PRs #163–#166 connect native Codex sessions to the existing Mission lifecycle, add the explicit
-`$sortie-dogs` skill and `codex init`, restore operation Reviewer correction within existing write
-scope, and reconcile native terminal observations without reviving completed work from delayed events.
-Codex settings and authentication remain separate from OpenCode. Real Ubuntu Codex acceptance
-evidence and host-specific limits remain in [the acceptance record](docs/codex-mission-acceptance-20261006.md);
-they are not a new all-host or performance benchmark. Release CLI proof observes actual OpenCode
-Worker startup and model identity only, not Mission completion. See [release notes](docs/release-v0.13.9.md).
-
-## v0.13.8 runtime updates
-
-PR #161 captures native validation bindings through the readable resolved target of a Windows
-directory junction instead of enumerating its failing logical alias. Logical evidence labels,
-link identity and complete target bytes remain bound; unchanged readable junctions retain their
-evidence identity. External-link/cycle handling, actual source/output freshness and Review remain
-unchanged. No new approval, permissions or model changes. The PR's isolated native Windows fixture
-reached independent Review PASS and accepted completion after resuming the same Mission, without
-rerunning its successful check; this does not claim original product recovery or uninterrupted
-single-turn completion. Release receipts: `_testenv/releases/0.13.8/`;
-see [release notes](docs/release-v0.13.8.md).
-
-## v0.13.7 runtime updates (retained)
-
-PRs #158/#159 prevent late write-only reports from falsely invalidating bound non-generating
-native checks, while preserving actual source/output freshness, failed checks and legacy recipes.
-Whole-root grants without a concrete inventory keep full-candidate freshness. Non-Git Mission roots
-can prepare independent Review from declared nested repositories/worktrees/plain outputs, including
-committed content; corrupt Git/HEAD errors remain visible. Operator performs authorized local recovery
-in the same turn without another approval, changed requirements or reset spending. Ubuntu's Git
-filesystem-boundary diagnostic is recognized alongside the standard non-repository error.
-The PR native observation reached Reviewer startup, not final acceptance. Release receipts:
-`_testenv/releases/0.13.7/`; see [release notes](docs/release-v0.13.7.md).
-
-## v0.13.6 runtime updates (retained)
-
-PR #156 references recorded formal PASS evidence in acceptance summaries instead of fetching
-the same successful unit's native history again for display. Missing/failed proof retains history
-diagnostics; historical evidence is not current freshness or acceptance. Identical external artifact
-inventories are shared only inside one snapshot refresh, never across later calls. Operator guidance
-proceeds to acceptance when current evidence covers the request and no concrete gap remains.
-Existing freshness/review guards, provider/model selection and cache-prefix machinery remain unchanged.
-No end-to-end speedup or live token/cache-hit improvement was measured. Release receipts:
-`_testenv/releases/0.13.6/`; see [release notes](docs/release-v0.13.6.md).
-
-## v0.13.5 runtime updates (retained)
-
-PR #154 preserves stable model instructions and appends changed host state at native history
-boundaries, with complete current-state reconstruction after compaction. Genuine Reviewer tools
-have deterministic read-only-first ordering; the real correction-permission transition still remains.
-The final candidate restores v3 behavioral review and combined assignment/findings while retaining
-exact instruction discovery, inherited Reviewer formal-command delivery, literal local shell-file
-scope reconciliation and explicit repository-root read/write scope support. New whole-project
-captures avoid bookkeeping-only invalidation; legacy evidence and real source/artifact freshness remain.
-
-The final pre-release cycle 11 Anko sample passed official local score 1 and selected public probes 9/9
-in 26m1.866s at estimated $1.41968316. It was faster but 15.762% costlier than the earlier v3 sample,
-not combined cost-preserving optimization or a general quality/speed guarantee. Release receipts:
-`_testenv/releases/0.13.5/`. See [release notes](docs/release-v0.13.5.md) and
-[candidate tradeoffs/failures](docs/cache-prefix-loop-20261003.md). Native Worker startup is not completion.
-
-## v0.13.4 runtime updates (retained)
-
-PR #152 restores same-session Coordinator/Operator implementation and formal validation through
-`plan_units(executor="self")`, `start_direct_unit` and `finish_direct_unit`. Known single-unit work can
-combine Mission start and planning; Luna Fast/max Worker routing remains the default. Full generated
-contracts remain visible when an explicit Read line range covers the file. Native background
-responsiveness remains: a launch acknowledgement or idle root is not Mission completion.
-
-The initial independent Reviewer can investigate, correct, formally validate, deliver and self-recheck
-continuously in its original native Task. Later Major/Medium findings accumulate in that same correction
-context. Author self-recheck remains `self-rechecked`, `independent=false`, never independent `PASS`.
-A different Reviewer is conditional on concrete residual Major risk; unresolved Major/Medium findings
-still block acceptance. Operator owns final comparison and receipt.
-
-Inherited compiler scratch no longer falsely invalidates broad-scope formal proof. Host-observed Git
-delivery, caller-setting review and test-composition guidance reduce avoidable detours. Saved host
-completion cards remain in tool history/UI, while outgoing V2 model/compaction context omits only their
-presentation body, retaining receipt and evidence identities.
-
-Two runs of the same fixed pre-release v8 package completed the original Anko task with official local
-score 1 (F2P 9/9, P2P 94/94) and selected public probes 9/9. Times were 26m49s and 25m55s, costs
-$1.47908048 and $1.54610816; each saved more than seven minutes versus the recorded v5 sample.
-These limited same-task observations do not establish general speedup or a new SWE-bench score.
-Release preflight, full tests, fixed-commit Windows CI and native Worker-start receipts are retained in
-`_testenv/releases/0.13.4/`; startup/model identity is not task completion. See the
-[release notes](docs/release-v0.13.4.md) and [quality-loop evidence](docs/nightly-quality-loop-20261002.md).
+- **Operator / Coordinator / Reviewer / Advisor:** GPT-6.1 Sol, `xhigh`.
+- **Worker / Scout:** Luna Fast, `max`.
+- **OpenCode:** `openai/gpt-6.1-sol#xhigh` and `openai/gpt-6-luna-fast#max`.
+- **Codex:** native `gpt-6.1-sol` / `xhigh`; the Luna-fast alias maps to native
+  `gpt-6-luna` / `max` with separate `serviceTier: "priority"` (Fast).
 
-## Mission workflow
+Explicit host model selections remain authoritative. OpenCode routing and Codex native settings
+are separate; unsupported native models fail rather than silently changing route. Codex Fast uses
+subscription limits faster than Standard. Subscription USD cost and remaining allowance are not
+inferred from token totals.
 
-Use Operator → Worker when one useful unit and its meaningful formal check are known;
-use Operator → Coordinator → Worker for actual discovery or decomposition:
-
-- `dog-operator` states a few requirements/negative constraints and owns user decisions and final acceptance.
-  The host saves the original user message verbatim.
-- Hidden `dogs-coordinator` owns investigation, unit declarations, Worker/Scout/Advisor/Reviewer dispatch,
-  in-request write-scope extensions, and corrections. It can read/search and run confirmation shell commands;
-  it can implement and formally validate directly in its own session, or delegate a unit to Worker.
-- `dog-worker-v010` implements a host-generated unit within its file/directory write scopes.
-  Investigation commands need no pre-registration; formal checks retain real host-recorded results.
-- High-risk changes require an initial independent Reviewer with read/search access. Low-risk skips
-  are explicit and recorded. Reviewer-owned corrections follow the self-recheck policy above.
-- Fast-lane can include high-risk single-unit work; it never implies a review skip.
-- Investigation, edits, formal checks and requested Git delivery stay in the same implementing child.
-  Explicit user ordering is retained; no routine plan-approval or commit-only handoff is needed.
-- Unit progress appears on the running Task without stopping Coordinator or prompting Operator.
-
-The `v010` Mission profile is serial by design; background responsiveness does not add parallel writers.
-The stable profile's Luna fabric and parallel integration path are not exposed in this profile. More agents are not a
-goal; preserving quality while reducing unnecessary expensive work is.
+## Measured results
 
-### SWE-bench evaluation
+### Codex end-to-end completion
 
-Official SWE-bench Lite `dev` results on the same 23 public instances:
-
-| Candidate / benchmark adapter | Official resolved | Empty patches | Run conditions | Details |
-| --- | ---: | ---: | --- | --- |
-| v0.10.6 candidate (`859c396`) | 4 / 23 (17.4%) | 5 | Four inference slots | [Handoff](docs/swebench-handoff-2026-09-21.md) |
-| Frozen v0.10.14 build | 6 / 23 (26.1%) | — | Four slots; $1.50/task | [Per-task results](docs/benchmark-v0.10.14-dev23.md) |
-| v0.12.8 (`e0f8cef` adapter) | 5 / 23 (21.7%) | 9 | Four slots; budget amended across two batches; 30-minute timeout | [Campaign](docs/swebench-v0128-dev23-2026-09-26.md) |
-| v0.12.8 (`84ccdf1` main adapter) | 6 / 23 (26.1%) | 2 | Fresh 23-task run; four slots; 30-minute timeout | [Main-integrated run](docs/swebench-main-84ccdf1-dev23-2026-09-26.md) |
-| v0.12.15 (`0c9690d`) | 5 / 23 (21.7%) | 3 | Fresh 23-task ext4 retry; four slots; 40-minute timeout | [Campaign](docs/swebench-v01215-dev23-2026-09-27.md) |
-| v0.12.16 (`9b05a34` release; `b1a6c0e` runner) | 4 / 23 (17.4%) | 5 | Fresh 23-task run; eight slots; 40-minute timeout | [Campaign](docs/swebench-v01216-dev23-2026-09-27.md) |
-| v0.12.19 (`24f5386` release; matched rerun) | 8 / 23 (34.8%) | 0 | Eight slots; effective $2/instance; 40-minute timeout; one inference timeout | [Official result and provenance](docs/benchmarks/swebench-v01220-operation-observability-2026-09-28.md) |
-| v0.12.20 (`628eb81` release) | 7 / 23 (30.4%) | 0 | Eight slots; effective $2/instance; 40-minute timeout | [Official result and caveat](docs/benchmarks/swebench-v01220-operation-observability-2026-09-28.md) |
-| v0.12.25 (`49eb1e4` release) | 7 / 23 (30.4%) | 0 | Eight slots; $2/instance; $30 total cap; 20-minute progress check / 40-minute hard maximum | [Comparison baseline](#v0131-dev23-2026-09-30) |
-| v0.13.1 (`d19e8be` release; 2026-09-30) | **8 / 23 (34.8%)** | 0 | Eight slots; $2/instance; $46 total cap; 20-minute progress check / 40-minute hard maximum; GPT-6.1 Sol + Luna Fast | [Run summary](#v0131-dev23-2026-09-30) |
-
-Every row has 23 submitted official predictions; an empty patch counts against
-the score, not as a missing evaluation. The v0.10.14 report does not separately
-summarize empty patches. The v0.12.15 row is the separately approved fresh
-run after an initial `/tmp` quota failure, not an additional score for that
-failed attempt. Inference completion is **not** official resolution.
-Budgets, runtime/adapter versions, and execution conditions changed between
-campaigns, so this table is a history of observed results, not a controlled
-head-to-head comparison or a general success-rate claim. The v0.10.14 run
-estimated $15.75 in model cost and a 15.2-minute median agent runtime.
-The v0.12.16 run is the first eight-slot inference score in this table;
-five runners timed out, and this does not establish a model-quality regression
-against runs with different concurrency and conditions.
-The v0.12.19 row is the corrected run with an effective $2 per-instance cap;
-an earlier v0.12.19 run scored 7/23 but had no effective per-instance cap and
-is not a same-condition comparison. The v0.12.20 run lost the
-`sqlfluff__sqlfluff-2419` resolution relative to the corrected run; this
-single run-to-run difference does not establish causation.
-
-Historical qualification references remain in [benchmark reference](docs/benchmark-reference.md).
+**Codex integration has completed real Missions on Ubuntu and Windows.** The Windows large-write
+Anko trial reached final Operator acceptance after Worker validation and same-Reviewer correction:
 
-#### v0.13.1 dev23 (2026-09-30)
+- SDK exit **0**, `accepted: true`, Mission **completed**.
+- **190/190** terminal command receipts; no unresolved command. Large generated-parser writes completed.
+- Separate local replay of official tests: reward **1**, F2P **9/9**, P2P **94/94**.
+- Elapsed **30m08.739s**, including cleanup; ChatGPT subscription monetary cost unavailable.
 
-One fresh pass@1 run and one official SWE-bench harness evaluation resolved
-**8/23**, versus **7/23** for v0.12.25. The new resolution was
-`pylint-dev__astroid-1333`; all seven previously resolved IDs were retained.
-Resolved by repository: marshmallow **2/2**, pvlib **0/5**, pydicom **2/5**,
-astroid **3/5**, pyvista **0/1**, sqlfluff **1/5**.
-
-- The scored row is the user-requested fresh run after a host restart. The
-  interrupted initial run is excluded from this score; the fresh run made one
-  attempt per instance with no inference retry.
-- The dataset revision (`6ec7bb89b9342f664a54a6e0a6ea6501d3437cc2`), public rows,
-  and all 23 official evaluation image IDs match the v0.12.25 run. Both used
-  `official-image-testbed` and sequential official scoring.
-- Operator/Coordinator/Reviewer/Advisor defaults changed to
-  `openai/gpt-6.1-sol#xhigh`. Actual task Workers remained
-  `openai/gpt-6-luna-fast#max`, observed across all 23 instances. The harness and
-  total budget also changed, so the extra resolution cannot be attributed to
-  the model change alone.
-- Inference ended with 20 normal completions, two timeouts
-  (`pvlib__pvlib-python-1154`, `sqlfluff__sqlfluff-1763`) and one agent failure
-  (`pvlib__pvlib-python-1854`). All patches, including stopped attempts, were
-  officially scored: 23 completed evaluations, zero empty patches and zero
-  official evaluation errors or infrastructure failures.
-- Known estimated inference cost: **$17.73**; separate unknown-usage hold:
-  **$1.98**, not counted as known expense. Inference wall time was about
-  **81 minutes**, followed by **7.2 minutes** of official scoring.
-- Fixed release commit: `d19e8be0d21180cc23ad2ae4b853d846a18e77bc`;
-  package SHA-256: `99300ceec0c3eee4fa1f984fed50d15d58b2ff4455df0041850ecd63514b0a51`.
-  OpenCode **2.0.20**, official harness **5.0.2**. Local evidence is retained in
-  `_testenv/swebench-v0131-dev23-20260930-r2/result-summary.json`; generated
-  predictions, databases and raw logs are not committed.
-
-## Mission tools
-
-1. `start_mission`: Operator supplies concise requirements; the host saves original messages. A known single unit
-   can include `unit` to combine start/planning and return its configured Worker task.
-2. `plan_units`: Operator or Coordinator supplies title, objective, file/directory scopes and formal checks. The host generates
-   IDs, handoff, manifest, proof mapping and the ready Worker task. `executor="self"` keeps execution in the
-   same controller session; `start_direct_unit` / `finish_direct_unit` retain observed formal-check freshness.
-3. `operator_next`: advance serial units. `expand_unit` reconciles required in-request outputs while
-   preserving the same Task. A reasoned `plan_units` correction or `retry_mission_unit` handles ordinary
-   unit recovery under the original requirements and cumulative budget.
-4. `review_mission`: generate the independent review packet from source, requirements and observed checks;
-   dispatch its Reviewer task for high-risk changes or record a low-risk skip.
-5. `repair_review`: record findings and continue correction in the running original Reviewer Task, or resume
-   that same native session. Later findings accumulate; run inherited checks/requested Git delivery and
-   explicitly report `SELF_RECHECKED` in that same Task. Legacy
-   `CORRECTION_READY` alone requires a same-author read-only fallback through `review_mission`.
-6. `submit_mission`: Coordinator returns a completion candidate, user-only decision, or proven external/scope/budget blocker.
-7. `complete_mission`: Operator compares the original request, source and evidence, then explicitly accepts.
-   Only a succeeded receipt authorizes DONE and the measured 🐾 return report.
-
-All tool names use the `sortie_v010_` prefix. Prior proposal/plan-repair tools remain in the compatibility
-implementation but are hidden from the normal Mission tool list. Operator/Coordinator handle in-request
-path reconciliation without a new user approval. Changes beyond the original requirements or cumulative
-budget return to Operator/user. `EVIDENCE_GAPS` is an advisory limitation, not Review `PASS` or an
-automatic extra review; failed or missing required checks still prevent acceptance.
-
-Durable profile state and hash-bound task references support restart and
-compaction recovery without reconstructing criteria from summary prose. Stale,
-foreign-root, or changed references are rejected. Requested `git add <paths>` and `git commit -m ...`
-use the actual source write scope, not a fabricated `.git/**` scope. An optional host-managed Git
-lifecycle also retains its branch, commit and post-commit boundaries. Neither mode grants arbitrary
-Git, force push, release or publication authority.
-
-### Progress and acceptance evidence
-
-`sortie_v010_operator_status` keeps original requests, formal command/exit/timing observations,
-review disposition and recorded delivery in a compact Mission view. `{ "view": "progress" }` exposes
-the current unit, completed/total units, host budget and next action; `{ "view": "full" }` or
-`details_ref` provides full snapshot diagnostics. Unknown clean state or a failed commit is not delivery
-success. Reading progress does not dispatch, retry or accept work; use native completion notifications
-instead of polling. Existing status reconciliation can recover a missed child terminal event.
+The run used native Codex models/threads with an **explicitly authorized SDK parent-host executor**.
+It does not claim native Windows sandbox repair, Docker-equivalent scoring or a hosted leaderboard
+submission. Its fixed development package preceded v0.13.11, which includes the changes; the
+published release was not rerun or assigned that score. These are completion observations, not a
+general speed or success-rate guarantee.
 
-## Configuration
+[Windows run and scoring](docs/codex-write-status-anko-20261007.md) ·
+[Windows models/shell evidence](docs/codex-windows-luna-fast-20261007.md) ·
+[Ubuntu acceptance](docs/codex-mission-acceptance-20261006.md)
 
-### Profile files and precedence
-
-The default package entry is the `v010` Mission profile:
-
-- Command: `/sortie-v010`
-- Primary agent: `dog-operator`
-- Project settings: `.opencode/sortie-dogs-v010.json`
-- Global settings: `~/.config/opencode/sortie-dogs-v010.json`
-- JSON environment override: `SORTIE_DOGS_V010_CONFIG`
-- Runtime state: `.sortie-dogs-v010/`
-- Installed asset marker: `.opencode/sortie-dogs-v010.version`
-
-For a global install, the marker is `<OpenCode config root>/sortie-dogs-v010.version`.
-
-Precedence is built-in defaults, global file, project file, environment JSON,
-then plugin factory options. Unknown properties or invalid types are rejected.
-Use external `v010` role names such as `dog-operator`, `dogs-coordinator`, and
-`dog-reviewer-v010` in `modelRouting`; do not also declare their stable aliases.
-
-Example `.opencode/sortie-dogs-v010.json`:
-
-```json
-{
-  "validationProfile": "balanced",
-  "readOnlyTools": ["my_mcp_search"],
-  "freeTierFallbackModels": ["opencode/deepseek-v4-flash-free"],
-  "modelRouting": {
-    "dog-operator": {
-      "preferred": { "model": "provider/model", "variant": "high" }
-    },
-    "dogs-coordinator": {
-      "preferred": { "model": "provider/model", "variant": "deep" }
-    }
-  },
-  "modelCatalog": {
-    "project": [
-      { "model": "provider/model", "variants": ["high", "deep"] }
-    ]
-  },
-  "continuation": {
-    "enabled": true,
-    "taskWatchdogMilliseconds": 300000
-  }
-}
-```
+### SWE-bench evaluation
 
-Declare only models and named variants the host actually provides. Sortie does
-not invent, probe, or translate variant names.
-
-### Settings reference
-
-- `readOnlyTools`: additional host-specific tools known not to mutate project
-  files. Values accumulate across configuration layers. Unknown tools are denied
-  in a bound worker session.
-- `modelRouting`: preferred and ordered fallback targets by external profile role.
-- `modelCatalog`: available `project` and `global` model/variant declarations.
-- `freeTierFallbackModels`: ordered global last-resort model IDs. Default:
-  `opencode/deepseek-v4-flash-free`; `[]` disables this fallback.
-- `dedicatedWorkerModel`: canonical stable serial target, default
-  `openai/gpt-6.1-sol` / `medium`. The Mission profile supplies its explicit
-  role routes below; do not infer its Worker route from this stable setting.
-- `consultation.strategy`: fixed advisor identity, optional `required`, and
-  positive `maxCallsPerCandidate`; default one call and not required.
-- `consultation.sourceReview`: risk-based review with `maxCallsPerCandidate`
-  default `1` and `maxArtifactBytes` default/maximum `30720`. Unavailable review
-  blocks only when review is required.
-- `continuation.enabled`: default `true`.
-- Automatic continuation has no turn-count ceiling. The legacy positive-integer
-  `continuation.maxAutoContinues` setting is accepted but ignored.
-- `continuation.taskWatchdogMilliseconds`: root inactivity while an implementation
-  Task is outstanding; default `300000`, valid range `10..1800000`.
-- `continuation.summarizeModel`: optional explicit compaction model; omission
-  reuses the latest observed root model.
-- `validationProfile`: `fast`, `balanced`, or `assurance`; default `balanced`.
-- `reflection`: enabled by default for `run`, `project` and `global` layers, with at most three
-  entries / 500 estimated tokens injected. Root Operator can use `sortie_v010_reflection` to retain
-  verified process causes/preventions for later turns and sessions; this is not model training.
-  Storage and managed blocks are separate from stable. Set `reflection.enabled` to `false` to disable.
-
-The Mission host owns handoff and manifest controls under
-`.sortie-dogs-v010/contracts/`. Do not create a legacy root
-`operation-manifest.json` for this profile and do not edit generated controls.
-Delete `.sortie-dogs-v010/` only when no Sortie run is active.
-
-### Validation policy
-
-`validationProfile` chooses supplementary non-canonical depth:
-
-- `fast`: static checks
-- `balanced`: targeted checks
-- `assurance`: related checks
-
-It does not replace meaningful declared formal checks or user/project-required broad validation.
-Batch related edits, run focused checks, then execute every declared formal check in order on the
-stable candidate. The implementing Worker or admitted correcting Reviewer runs those commands;
-canonical/full-suite `owner=coordinator` is evidence accounting, not a requirement for root execution.
-
-Keep required broad checks for the final integrated candidate. Reuse valid unchanged evidence only
-when the contract permits; identity includes candidate, command, environment, scope and owner.
-Required repeated occurrences retain their own execution identities and cannot be skipped as duplicates.
-Native commands, actual working directory, exit, duration and saved source bindings establish freshness.
-Diagnostics do not substitute for formal proof. Repeat checks when changes, failures or freshness require
-it, rather than solely because a Worker changed or documentation was edited.
-
-### Default routes
-
-- `dog-operator`: `openai/gpt-6.1-sol` / `xhigh`
-- `dogs-coordinator`: `openai/gpt-6.1-sol` / `xhigh`
-- `dog-worker-v010`: `openai/gpt-6-luna-fast` / `max`
-- `dog-scout-v010`: `openai/gpt-6-luna-fast` / `max`
-- `dog-reviewer-v010`: `openai/gpt-6.1-sol` / `xhigh`
-- `dog-advisor-v010`: `openai/gpt-6.1-sol` / `xhigh`
-
-An explicit model and variant selected in OpenCode remains authoritative for that
-session. Child role defaults fill absent native settings and may be overridden by
-valid profile routing. Review never silently inherits the implementation model.
-
-## Stable compatibility profile
-
-The earlier parallel-capable runtime remains available explicitly:
+**SWE-bench Lite test300: 170/300 (56.67%)**, fixed **v0.12.24**, one pass@1 campaign.
+Nine empty patches, zero official evaluation errors, frozen predictions and trajectories for all
+300 instances. Workers ran Luna Fast/max; management/review ran GPT-6 Sol/xhigh.
+This is a system result, **not** a Luna-only comparison or Verified/full SWE-bench score.
 
-```sh
-npx sortie-dogs init . --profile stable
-```
+Confirmed inference expense **$162.99**; separate unknown-pricing budget hold **$34.60** is not
+known expense. Official local evaluation and leaderboard acceptance are separate.
 
-Load it through a project bridge instead of the default package plugin entry:
+[Technical report](docs/benchmarks/swebench-lite-v01224-test300-2026-09-29.md) ·
+[Public predictions/logs/trajectories](https://github.com/zufall-upon/sortie-dogs-swebench-lite-20260929) ·
+[Result history and conditions](docs/benchmarks/swebench-lite-history.md)
 
-```ts
-export { SortieDogsPlugin } from "sortie-dogs/plugin/stable";
-```
+#### v0.13.1 dev23 (2026-09-30)
 
-The stable profile uses `/sortie`, `dog-coordinator`,
-`.opencode/sortie-dogs.json`, `SORTIE_DOGS_CONFIG`, and `.sortie-dogs/`. Do not
-register stable and `v010` from the same package installation path in one host.
+**8/23 (34.8%)**, versus v0.12.25's 7/23; all seven prior resolutions retained, no empty patches or
+official evaluation errors. [Fixed conditions, stopped-run accounting and costs](docs/benchmarks/swebench-lite-history.md#v0131-dev23-2026-09-30).
 
-## Global availability
+All scores belong to their fixed candidates, **not v0.13.13**. SWE-bench remains an optional
+measurement, not a release gate. Different budgets, models and conditions are not controlled comparisons.
 
-Project-local installation is recommended. To expose the current Mission assets globally:
+## Configuration
 
-```sh
-npm install --global sortie-dogs@0.13.10
-sortie-dogs init --global --profile v010
-```
+**OpenCode:** project settings `.opencode/sortie-dogs-v010.json`, global settings
+`~/.config/opencode/sortie-dogs-v010.json`, JSON override `SORTIE_DOGS_V010_CONFIG`.
+The default Mission state is `.sortie-dogs-v010/`; generated controls are host-owned.
+[Precedence, settings, global installation and stable compatibility](docs/configuration.md).
 
-Global initialization registers the package or reuses an existing local V2 bridge, and sets subagent
-depth to at least two, preserving unrelated settings and the default agent.
+**Codex:** use `codex mission --help` or `CodexMissionSession`. Optional `--model`, `--effort`,
+`--permissions`, `--trusted-pwsh` and `--resume` are explicit host choices, not OpenCode settings.
+[SDK and recovery reference](docs/codex.md).
 
-An existing `<OpenCode config root>/plugins/sortie-dogs/index.js` bridge importing `sortie-dogs/server`
-can resolve a **separate dependency** under that config root. Updating npm-global alone does not update
-it. For that layout, also install the same release at the actual config root, then rerun global init:
+## Updates and removal
 
-```sh
-npm install --prefix "$HOME/.config/opencode" sortie-dogs@0.13.10
-sortie-dogs init --global --profile v010
-```
+Install the desired package version and rerun **your host's** initializer:
 
-The command shows the default config root; use your actual root if overridden. For a configured npm
-package entry, OpenCode V2 also provides `opencode plugin list` / `opencode plugin update`; exact
-version pins require an explicit version change. Completely restart OpenCode after updating, then
-check the installed package, asset marker and loaded plugin version.
+- Codex: `npx --no-install sortie-dogs codex init .`, then restart/open a new Codex chat.
+- OpenCode: `npx --no-install sortie-dogs init .`, then completely restart OpenCode.
 
-## Updates and removal
+An OpenCode config-local bridge may resolve a separate package; npm-global update alone does not
+update it. [Global/update instructions](docs/configuration.md#global-availability).
+The current Mission asset marker is `0.13.13-mission-context-v1`; the unchanged Codex skill marker is
+`0.13.11-codex-skill-v2`. Installed markers alone do not prove a running host reloaded the new version.
 
-For project-local updates, replace the dependency, rerun initialization and completely restart OpenCode:
+There is no supported uninstall command. Remove the dependency and only known Sortie-owned paths;
+follow the [manual removal guide](docs/uninstall.md), never delete the entire `.opencode` directory.
 
-```sh
-npm install --save-dev sortie-dogs@latest
-npx sortie-dogs init .
-```
+## Documentation
 
-`init` is idempotent. It updates recognized Sortie-owned assets, records the asset
-version, preserves user configuration, and stops safely on unknown ownership or
-conflicting files.
+- **Usage:** [Codex guide](docs/codex.md) · [OpenCode configuration](docs/configuration.md) ·
+  [日本語 OpenCodeガイド](docs/guide-ja.md) · [简体中文 OpenCode指南](docs/guide-zh-CN.md).
+- **Design:** [Worker instructions](docs/worker-instruction-design.md) ·
+  [Quality-first review](docs/quality-first-review.md) · [Operation efficiency](docs/mission-operation-efficiency.md).
+- **Evidence:** [Codex completion](docs/codex-write-status-anko-20261007.md) ·
+  [SWE-bench history](docs/benchmarks/swebench-lite-history.md) · [Historical local case study](docs/benchmark-reference.md).
+- **Development:** [Testing](docs/testing.md) · [CLI testing](docs/cli-testing.md) ·
+  [Windows tests](docs/windows-tests.md) · [Release routine](docs/release-batch.md).
+- **Releases:** [v0.13.13](docs/release-v0.13.13.md) · [v0.13.12](docs/release-v0.13.12.md) · [v0.13.11](docs/release-v0.13.11.md) · [v0.13.10](docs/release-v0.13.10.md) ·
+  [v0.13.9](docs/release-v0.13.9.md) · [All GitHub releases](https://github.com/zufall-upon/Sortie-dogs/releases).
 
-Align any exact version pin or separate bridge dependency with the intended release too. An installed
-marker of `0.13.10-codex-windows-v1` identifies the assets; it does not prove an already-running
-OpenCode process has reloaded the plugin.
+## Community
 
-There is no supported uninstall command. Remove the npm dependency separately,
-then follow the [safe manual removal guide](docs/uninstall.md). Delete only known
-Sortie-owned paths; never remove the whole `.opencode` directory or use broad
-wildcards.
+[Contributing](CONTRIBUTING.md) · [Code of conduct](CODE_OF_CONDUCT.md) ·
+[Security policy / private reports](SECURITY.md) · [Accessibility](ACCESSIBILITY.md) ·
+[Report a bug or propose a feature](https://github.com/zufall-upon/Sortie-dogs/issues/new/choose).
 
-Maintainers: the [release batch guide](docs/release-batch.md) covers fixed-tarball
-validation, global application, GitHub publication, and manual npm publication.
+Licensed under [MIT](LICENSE).