sortie-dogs 0.14.5 → 0.14.7

package/.agents/skills/sortie-dogs/SKILL.md

@@ -32,7 +32,7 @@ Use the natural-language Mission route on POSIX and Windows:
 
 Forward `--model`, `--effort`, or `--permissions` only when the user selected those values. Never silently widen the native permission profile. If an existing Mission is reported, resume only its exact root with `--resume`; do not launch replacement work.
 
-Leave `--model` unset to retain role-specific defaults: Sol 6.1 Operator/Coordinator/Reviewer and Luna-fast Worker. The Codex adapter maps the shared Luna-fast alias to native `gpt-6-luna` plus the `priority` (Fast) service tier; it does not alter OpenCode defaults. Fast consumes subscription limits faster than Standard; no metered API fallback is allowed.
+Leave `--model` unset to retain role-specific defaults: Sol 6.1 Operator/Coordinator/Reviewer and ordinary Terra 6.1 Standard/max Worker (no Fast/priority). Scout and dedicated Luna routes remain unchanged. The Codex adapter retains the shared Luna-fast alias mapping to native `gpt-6-luna` plus the `priority` (Fast) service tier. Fast consumes subscription limits faster than Standard; no metered API fallback is allowed.
 
 On Windows the same Mission command resolves existing PowerShell 7 automatically. Add `--trusted-pwsh <absolute-pwsh.exe>` only when an exact existing executable needs to be selected; do not add a separate shell preflight or approval step.
 

package/README.md

@@ -19,8 +19,8 @@ evidence-backed result, not more agents for their own sake.
 [![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)
 
-**Current release: [v0.14.5](https://github.com/zufall-upon/Sortie-dogs/releases/tag/v0.14.5)**
-· [Release notes](docs/release-v0.14.5.md) · [Execution vision](docs/v0.14.0-execution-vision.md)
+**Current release: [v0.14.7](https://github.com/zufall-upon/Sortie-dogs/releases/tag/v0.14.7)**
+· [Release notes](docs/release-v0.14.7.md) · [Execution vision](docs/v0.14.0-execution-vision.md)
 
 [Quick start](#quick-start) · [Workflow](#mission-workflow) · [Models](#default-models) ·
 [Measured results](#measured-results) · [Configuration](#configuration) · [Documentation](#documentation)
@@ -57,7 +57,7 @@ Use an existing Codex CLI signed in with **ChatGPT**. On Windows, PowerShell **7
 available; POSIX uses bash. Sortie does not install Codex, start login or use a metered API fallback.
 
 ```sh
-npm install --save-dev sortie-dogs@0.14.5
+npm install --save-dev sortie-dogs@0.14.7
 npx --no-install sortie-dogs codex init .
 ```
 
@@ -86,7 +86,7 @@ bridge. [Codex guide: SDK, permissions, progress, resume and manifest route](doc
 Use **OpenCode V2**. In the target project:
 
 ```sh
-npm install --save-dev sortie-dogs@0.14.5
+npm install --save-dev sortie-dogs@0.14.7
 npx --no-install sortie-dogs init .
 ```
 
@@ -144,9 +144,12 @@ one. Required success checks, missing commands and active/unstarted operations s
 ## Default models
 
 - **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
+- **Worker:** GPT-6.1 Terra Standard, `max` (no Fast/priority).
+- **Scout / dedicated Luna Worker:** Luna Fast, `max`, unchanged.
+- **OpenCode:** ordinary Worker `openai/gpt-6.1-terra#max`; management/review
+  `openai/gpt-6.1-sol#xhigh`.
+- **Codex:** ordinary Worker native `gpt-6.1-terra` / `max`, Standard; management/review
+  `gpt-6.1-sol` / `xhigh`. The retained Luna-fast alias maps to native
   `gpt-6-luna` / `max` with separate `serviceTier: "priority"` (Fast).
 
 Explicit host model selections remain authoritative. OpenCode routing and Codex native settings
@@ -242,8 +245,8 @@ Install the desired package version and rerun **your host's** initializer:
 
 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 `0.14.5-unborn-proof-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.
+The current Mission asset marker is `0.14.7-visible-recovery-v1`; the Codex skill marker is
+`0.14.6-codex-skill-v1`. Installed markers alone do not prove a running host reloaded the new version.
 
 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.
@@ -258,9 +261,14 @@ follow the [manual removal guide](docs/uninstall.md), never delete the entire `.
   [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:** [v0.14.5](docs/release-v0.14.5.md) · [v0.14.4](docs/release-v0.14.4.md) · [v0.14.3](docs/release-v0.14.3.md) · [v0.14.2](docs/release-v0.14.2.md) · [v0.14.1](docs/release-v0.14.1.md) · [v0.14.0](docs/release-v0.14.0.md) · [v0.13.22](docs/release-v0.13.22.md) · [v0.13.21](docs/release-v0.13.21.md) · [v0.13.20](docs/release-v0.13.20.md) · [v0.13.19](docs/release-v0.13.19.md) · [v0.13.18](docs/release-v0.13.18.md) · [v0.13.17](docs/release-v0.13.17.md) · [v0.13.16](docs/release-v0.13.16.md) · [v0.13.15](docs/release-v0.13.15.md) · [v0.13.14](docs/release-v0.13.14.md) · [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) ·
+- **Releases:** [v0.14.7](docs/release-v0.14.7.md) · [v0.14.6](docs/release-v0.14.6.md) · [v0.14.5](docs/release-v0.14.5.md) · [v0.14.4](docs/release-v0.14.4.md) · [v0.14.3](docs/release-v0.14.3.md) · [v0.14.2](docs/release-v0.14.2.md) · [v0.14.1](docs/release-v0.14.1.md) · [v0.14.0](docs/release-v0.14.0.md) · [v0.13.22](docs/release-v0.13.22.md) · [v0.13.21](docs/release-v0.13.21.md) · [v0.13.20](docs/release-v0.13.20.md) · [v0.13.19](docs/release-v0.13.19.md) · [v0.13.18](docs/release-v0.13.18.md) · [v0.13.17](docs/release-v0.13.17.md) · [v0.13.16](docs/release-v0.13.16.md) · [v0.13.15](docs/release-v0.13.15.md) · [v0.13.14](docs/release-v0.13.14.md) · [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).
 
+## Support
+
+[GitHub Sponsors](https://github.com/sponsors/zufall-upon) — sponsorships become available
+once the maintainer's Sponsors profile is approved and published.
+
 ## Community
 
 [Contributing](CONTRIBUTING.md) · [Code of conduct](CODE_OF_CONDUCT.md) ·

package/dist/asset-version.d.ts

@@ -3,6 +3,6 @@
  * installed project marker without importing every asset body.
  */
 export declare const RUNTIME_ASSET_VERSION = "0.3.89-completion-proof-v1";
-export declare const V010_RUNTIME_ASSET_VERSION = "0.14.5-unborn-proof-v1";
-export declare const CODEX_SKILL_ASSET_VERSION = "0.13.11-codex-skill-v2";
+export declare const V010_RUNTIME_ASSET_VERSION = "0.14.7-visible-recovery-v1";
+export declare const CODEX_SKILL_ASSET_VERSION = "0.14.6-codex-skill-v1";
 export type RuntimeAssetVersion = typeof RUNTIME_ASSET_VERSION | typeof V010_RUNTIME_ASSET_VERSION | typeof CODEX_SKILL_ASSET_VERSION;

package/dist/asset-version.js

@@ -3,5 +3,5 @@
  * installed project marker without importing every asset body.
  */
 export const RUNTIME_ASSET_VERSION = "0.3.89-completion-proof-v1";
-export const V010_RUNTIME_ASSET_VERSION = "0.14.5-unborn-proof-v1";
-export const CODEX_SKILL_ASSET_VERSION = "0.13.11-codex-skill-v2";
+export const V010_RUNTIME_ASSET_VERSION = "0.14.7-visible-recovery-v1";
+export const CODEX_SKILL_ASSET_VERSION = "0.14.6-codex-skill-v1";

package/dist/cli/codex-mission.d.ts

@@ -1,5 +1,5 @@
 import type { CodexTurnEvent } from "../codex/app-server.js";
-export declare const CODEX_MISSION_USAGE = "Usage: sortie-dogs codex mission --prompt <text>\n  [--resume <root-thread-id>] [--project-root <path>] [--executable <codex>] [--model <model>] [--effort <effort>] [--permissions <native-profile>] [--trusted-pwsh <absolute-pwsh.exe>]\n\nRuns the existing Mission Operator, Coordinator, Worker and Reviewer through Codex.\n--model and --effort explicitly override all roles; otherwise packaged role defaults apply.\n--permissions explicitly selects an existing native profile; omission retains native defaults.\nWindows uses existing PowerShell 7; --trusted-pwsh pins its absolute executable path.\nThe packaged Luna-fast worker maps to native gpt-6-luna with the priority (Fast) service tier.\nUses existing ChatGPT authentication and native Codex permissions.";
+export declare const CODEX_MISSION_USAGE = "Usage: sortie-dogs codex mission --prompt <text>\n  [--resume <root-thread-id>] [--project-root <path>] [--executable <codex>] [--model <model>] [--effort <effort>] [--permissions <native-profile>] [--trusted-pwsh <absolute-pwsh.exe>]\n\nRuns the existing Mission Operator, Coordinator, Worker and Reviewer through Codex.\n--model and --effort explicitly override all roles; otherwise packaged role defaults apply.\n--permissions explicitly selects an existing native profile; omission retains native defaults.\nWindows uses existing PowerShell 7; --trusted-pwsh pins its absolute executable path.\nThe packaged ordinary worker uses native gpt-6.1-terra/max, Standard (no Fast/priority).\nThe retained Luna-fast alias maps to native gpt-6-luna with the priority (Fast) service tier.\nUses existing ChatGPT authentication and native Codex permissions.";
 export declare function runCodexMissionCommand(argv: readonly string[]): Promise<number>;
 /** Bounded user-facing progress; do not dump control packets or hidden reasoning. */
 export declare function codexMissionProgress(event: CodexTurnEvent & {

package/dist/cli/codex-mission.js

@@ -6,7 +6,8 @@ Runs the existing Mission Operator, Coordinator, Worker and Reviewer through Cod
 --model and --effort explicitly override all roles; otherwise packaged role defaults apply.
 --permissions explicitly selects an existing native profile; omission retains native defaults.
 Windows uses existing PowerShell 7; --trusted-pwsh pins its absolute executable path.
-The packaged Luna-fast worker maps to native gpt-6-luna with the priority (Fast) service tier.
+The packaged ordinary worker uses native gpt-6.1-terra/max, Standard (no Fast/priority).
+The retained Luna-fast alias maps to native gpt-6-luna with the priority (Fast) service tier.
 Uses existing ChatGPT authentication and native Codex permissions.`;
 export async function runCodexMissionCommand(argv) {
     if (argv.length === 1 && argv[0] === "--help") {

package/dist/core/mission-context.d.ts

@@ -1,3 +1,12 @@
+export interface MissionAttachment {
+    index: number;
+    part_id?: string;
+    mime?: string;
+    filename?: string;
+    url: string;
+    path?: string;
+    sha256?: string;
+}
 export interface PublicMissionContext {
     id: string;
     role: "user" | "assistant";
@@ -12,6 +21,7 @@ export interface PublicMissionContext {
         exit?: number;
         time?: unknown;
     }[];
+    attachments?: MissionAttachment[];
 }
 export interface MissionContextSource {
     session_id: string;
@@ -20,6 +30,9 @@ export interface MissionContextSource {
     entries: number;
     observed_at: string;
     history_available: boolean;
+    attachments?: (MissionAttachment & {
+        message_id: string;
+    })[];
 }
 export interface MissionContinuity {
     sources: MissionContextSource[];
@@ -70,4 +83,7 @@ export declare function missionContextView(source: MissionContextSource, entries
     entries: number;
     observed_at: string;
     history_available: boolean;
+    attachments?: (MissionAttachment & {
+        message_id: string;
+    })[];
 };

package/dist/core/mission-context.js

@@ -1,6 +1,7 @@
 import { createHash } from "node:crypto";
 import { mkdir, readFile, writeFile } from "node:fs/promises";
 import { join } from "node:path";
+import { fileURLToPath, pathToFileURL } from "node:url";
 const record = (value) => value !== null && typeof value === "object" && !Array.isArray(value);
 /** Public task data only. Keep exact text/results in a lazy snapshot, not private reasoning. */
 export function publicMissionContext(messages) {
@@ -11,6 +12,11 @@ export function publicMissionContext(messages) {
         const parts = message.parts.filter(record);
         const text = parts.filter(part => part.type === "text" && part.synthetic !== true && typeof part.text === "string")
             .map(part => part.text).join("\n");
+        const attachments = info.role === "user" ? parts.filter(part => part.type === "file" && typeof part.url === "string")
+            .map((part, index) => ({ index, url: part.url,
+            ...(typeof part.id === "string" ? { part_id: part.id } : {}),
+            ...(typeof part.mime === "string" ? { mime: part.mime } : {}),
+            ...(typeof part.filename === "string" ? { filename: part.filename } : {}) })) : [];
         const tools = parts.flatMap(part => {
             if (part.type !== "tool" || typeof part.tool !== "string" || !record(part.state))
                 return [];
@@ -23,17 +29,47 @@ export function publicMissionContext(messages) {
                     ...(typeof metadata.exit === "number" ? { exit: metadata.exit } : {}),
                     ...(part.time !== undefined ? { time: part.time } : {}) }];
         });
-        return text.trim() || tools.length ? [{ id: info.id, role: info.role, text, tools }] : [];
+        return text.trim() || tools.length || attachments.length ? [{ id: info.id, role: info.role, text, tools,
+                ...(attachments.length ? { attachments } : {}) }] : [];
     });
 }
+/** Store inline bytes once, not megabytes of base64 in every prompt/contract. No network fetch or new permission. */
+async function retainAttachment(directory, attachment) {
+    if (!attachment.url.startsWith("data:")) {
+        try {
+            return attachment.url.startsWith("file:") ? { ...attachment, path: fileURLToPath(attachment.url) } : attachment;
+        }
+        catch {
+            return attachment;
+        } // Keep a host-specific URI usable on its owning host, not a new admission gate.
+    }
+    const comma = attachment.url.indexOf(",");
+    if (comma < 0)
+        return attachment;
+    const header = attachment.url.slice(0, comma), data = attachment.url.slice(comma + 1);
+    const bytes = /;base64$/i.test(header) ? Buffer.from(data, "base64") : Buffer.from(decodeURIComponent(data));
+    const sha256 = createHash("sha256").update(bytes).digest("hex");
+    const extension = { "image/png": ".png", "image/jpeg": ".jpg", "image/gif": ".gif", "image/webp": ".webp",
+        "application/pdf": ".pdf", "text/plain": ".txt" }[attachment.mime ?? ""] ?? ".bin";
+    const path = join(directory, "attachments", `${sha256}${extension}`);
+    await mkdir(join(directory, "attachments"), { recursive: true });
+    await writeFile(path, bytes, { flag: "wx" }).catch(error => {
+        if (error.code !== "EEXIST")
+            throw error;
+    });
+    return { ...attachment, url: pathToFileURL(path).href, path, sha256 };
+}
 /** Preserve earlier available entries when the host later returns a compacted/partial history. */
 export async function retainMissionContext(directory, sessionID, messages, previous) {
     const current = publicMissionContext(messages);
     const retained = previous ? await readFile(previous.path, "utf8")
         .then(content => JSON.parse(content).entries).catch(() => undefined) : undefined;
     const merged = new Map((retained ?? []).map(entry => [entry.id, entry]));
-    for (const entry of current)
-        merged.set(entry.id, entry);
+    for (const entry of current) {
+        const attachments = entry.attachments ? await Promise.all(entry.attachments.map(file => retainAttachment(directory, file)))
+            : merged.get(entry.id)?.attachments;
+        merged.set(entry.id, { ...entry, ...(attachments?.length ? { attachments } : {}) });
+    }
     const entries = [...merged.values()];
     const content = JSON.stringify({ session_id: sessionID, entries }, null, 2) + "\n";
     const sha256 = createHash("sha256").update(content).digest("hex");
@@ -42,8 +78,10 @@ export async function retainMissionContext(directory, sessionID, messages, previ
         await mkdir(directory, { recursive: true });
         await writeFile(path, content);
     }
+    const attachments = entries.flatMap(entry => (entry.attachments ?? []).map(file => ({ message_id: entry.id, ...file })));
     return { source: { session_id: sessionID, path, sha256, entries: entries.length,
-            observed_at: new Date().toISOString(), history_available: current.length > 0 }, entries };
+            observed_at: new Date().toISOString(), history_available: current.length > 0,
+            ...(attachments.length ? { attachments } : {}) }, entries };
 }
 /** Small inline view, with explicit pointers to every omitted byte in the exact snapshot. */
 export function missionContextView(source, entries, suppliedRequestIDs = []) {

package/dist/core/operator-mission.js

@@ -338,7 +338,7 @@ export class OperatorMissionRuntime {
                 corrections: (state.corrections ?? []).map(({ initialPrompt: _prompt, inlineReview: _inline, ...correction }) => correction),
                 retained_review: state.review ? { run_id: state.review.runID, verdict: state.review.verdict,
                     mode: state.review.mode, author: state.review.child, result: state.review.result, self_recheck: state.review.selfRecheck } : null,
-                guidance: "Current original_requests/requirements govern; older plans remain historical. Continue from known decisions, environment, completed work, retained findings and failures. A new plan or author does not resolve findings or turn self-recheck into independent approval. Read snapshot JSON entries by message/call IDs only for a relevant gap. Recheck mutable facts when needed or conflicting; do not repeat unchanged setup/investigation/checks merely because ownership changed. Successful evidence remains with its actual author/call and input bindings; historical results are not new authority." } };
+                guidance: "Current original_requests/requirements govern; older plans remain historical. Original attachments are in sources[].attachments, ordered within each message; read their path/URL when relevant, not a prose substitute. Message IDs and indices preserve provenance, not an inferred mapping to user-assigned image numbers. Continue from known decisions, environment, completed work, retained findings and failures. A new plan or author does not resolve findings or turn self-recheck into independent approval. Read snapshot JSON entries by message/call IDs only for a relevant gap. Recheck mutable facts when needed or conflicting; do not repeat unchanged setup/investigation/checks merely because ownership changed. Successful evidence remains with its actual author/call and input bindings; historical results are not new authority." } };
     }
     async load(file) {
         try {
@@ -870,6 +870,7 @@ export async function missionAcceptanceSummary(mission, run, operators, observe)
         }
     })));
     return { original_requests: mission.requests, requirements: mission.requirements,
+        original_attachments: (mission.continuity?.sources ?? []).flatMap(source => (source.attachments ?? []).map(file => ({ session_id: source.session_id, ...file }))),
         operation: missionOperationSummary(mission),
         history: { status: missingAnchors.length ? "unavailable" : history.status,
             ...(history.reason ? { reason: history.reason } : missingAnchors.length ? { reason: "accepted-handoff-anchor-unavailable" } : {}),
@@ -887,7 +888,7 @@ export async function missionAcceptanceSummary(mission, run, operators, observe)
             observation: mission.deliveryObservation && mission.deliveryObservation.run_id === current?.runID ? mission.deliveryObservation : null,
             observation_source: "persisted operator Git lifecycle and formal validation records; no new Git inspection",
             clean: mission.deliveryObservation && mission.deliveryObservation.run_id === current?.runID ? mission.deliveryObservation.clean : "not independently observed by this projection" },
-        interpretation: "Compare original requests with the submitted candidate and actual evidence. Historical PASS is not current PASS. Inspect concrete gaps, not routine archive searches or full source rereads. Existing completion and Review guards still apply." };
+        interpretation: "Compare original requests (including relevant original attachments) with the submitted candidate and what the actual evidence exercises. Functional PASS, Review PASS, or an honestly recorded unverified limitation does not establish a required visual comparison. If a mandatory outcome remains unmet, continue authorized retrieval, inspection or correction without waiting for the user. Optional caveats, valid alternative checks and reusable evidence need not block acceptance; do not reject on the word unverified alone or repeat all checks yourself. No new approval, proof file, extra Reviewer, universal screenshots or pixel-exact matching is required. Historical PASS is not current PASS. Inspect concrete gaps, not routine archive searches or full source rereads. Existing completion and Review guards still apply." };
 }
 export function missionReviewIndependent(mission, child) {
     return child !== undefined && !mission.sourceAuthors?.includes(child) &&

package/dist/plugin/index.js

@@ -8488,12 +8488,16 @@ export const SortieDogsPlugin = async (input, options) => {
             return settled;
         },
         releaseDirectUnit: async (actor) => {
-            await releaseWriteGate(actor);
+            const released = JSON.parse(await releaseWriteGate(actor));
+            if (released.status === "denied")
+                throw new Error(`mission-direct-release:${released.reason}`);
             releaseSessionEnforcement(actor);
         },
         assertDirectUnitQuiescent: async (actor) => {
             if (activeSessions.get(actor)?.inFlightCalls.size)
                 throw new Error("mission-direct-tools-running");
+            if (bindingOperations.has(actor))
+                throw new Error("mission-direct-binding-in-flight");
         },
         failDirectUnit: async (root, actor) => {
             const direct = await input.runtimeBridge?.directMissionExecution?.(actor);

package/dist/plugin/model-cost.d.ts

@@ -1,9 +1,9 @@
 export declare const MODEL_COST_PRICING_SNAPSHOT: {
-    readonly checkedAt: "2026-09-30";
+    readonly checkedAt: "2026-10-27";
     readonly currency: "USD";
     readonly unit: "1M tokens";
-    readonly sources: readonly ["https://developers.openai.com/api/docs/pricing", "https://developers.openai.com/api/docs/models/gpt-6-astra", "https://developers.openai.com/api/docs/models/gpt-6-sol", "https://developers.openai.com/api/docs/models/gpt-6.1-sol", "https://developers.openai.com/api/docs/models/gpt-6-luna", "https://developers.openai.com/api/docs/guides/fast-mode", "https://developers.openai.com/api/docs/models/gpt-5.6-sol", "https://developers.openai.com/api/docs/models/gpt-5.6-terra", "https://developers.openai.com/api/docs/models/gpt-5.6-luna", "https://platform.claude.com/docs/en/about-claude/pricing"];
-    readonly assumptions: readonly ["OpenAI Standard pricing; requests above 272,000 total input/cache tokens use 2x input/cache and 1.5x output pricing.", "Anthropic Standard pricing with the 5-minute cache-write rate; Batch, Flex, regional, tool, and subscription charges are excluded.", "variant and serviceTier are separate; model variants do not alter the selected Standard token price.", "GPT-6 Sol, GPT-6.1 Sol, and GPT-6 Luna use their official Standard schedules; the OpenCode gpt-6-luna-fast selector sends API model gpt-6-luna with priority service tier.", "For GPT-6 Sol, GPT-6.1 Sol, and GPT-6 Luna, Fast mode (service_tier fast or priority, including the gpt-6.1-sol-fast selector) multiplies applicable Standard rates by 2, including cache and long-context multipliers.", "gpt-5.6-luna-fast is retained only as a compatibility price for historical host catalog entries; it has no published model page and is exactly twice the GPT-5.6 Luna Standard schedule."];
+    readonly sources: readonly ["https://developers.openai.com/api/docs/pricing", "https://developers.openai.com/api/docs/models/gpt-6-astra", "https://developers.openai.com/api/docs/models/gpt-6-sol", "https://developers.openai.com/api/docs/models/gpt-6.1-sol", "https://developers.openai.com/api/docs/models/gpt-6.1-terra", "https://developers.openai.com/api/docs/models/gpt-6-luna", "https://developers.openai.com/api/docs/guides/fast-mode", "https://developers.openai.com/api/docs/models/gpt-5.6-sol", "https://developers.openai.com/api/docs/models/gpt-5.6-terra", "https://developers.openai.com/api/docs/models/gpt-5.6-luna", "https://platform.claude.com/docs/en/about-claude/pricing"];
+    readonly assumptions: readonly ["OpenAI Standard pricing; requests above 272,000 total input/cache tokens use 2x input/cache and 1.5x output pricing.", "Anthropic Standard pricing with the 5-minute cache-write rate; Batch, Flex, regional, tool, and subscription charges are excluded.", "variant and serviceTier are separate; model variants do not alter the selected Standard token price.", "GPT-6.1 Terra Standard uses $1 input, $0.05 cache read, $1.25 30-minute cache write and $5 output per 1M tokens; subscription usage/limits are separate from this API-equivalent estimate.", "GPT-6 Sol, GPT-6.1 Sol, and GPT-6 Luna use their official Standard schedules; the OpenCode gpt-6-luna-fast selector sends API model gpt-6-luna with priority service tier.", "For GPT-6 Sol, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna, Fast mode (service_tier fast or priority, including the gpt-6.1-sol-fast selector) multiplies applicable Standard rates by 2, including cache and long-context multipliers.", "gpt-5.6-luna-fast is retained only as a compatibility price for historical host catalog entries; it has no published model page and is exactly twice the GPT-5.6 Luna Standard schedule."];
 };
 export interface ModelCostUsage {
     readonly providerID: string | undefined;

package/dist/plugin/model-cost.js

@@ -1,5 +1,5 @@
 export const MODEL_COST_PRICING_SNAPSHOT = {
-    checkedAt: "2026-09-30",
+    checkedAt: "2026-10-27",
     currency: "USD",
     unit: "1M tokens",
     sources: [
@@ -7,6 +7,7 @@ export const MODEL_COST_PRICING_SNAPSHOT = {
         "https://developers.openai.com/api/docs/models/gpt-6-astra",
         "https://developers.openai.com/api/docs/models/gpt-6-sol",
         "https://developers.openai.com/api/docs/models/gpt-6.1-sol",
+        "https://developers.openai.com/api/docs/models/gpt-6.1-terra",
         "https://developers.openai.com/api/docs/models/gpt-6-luna",
         "https://developers.openai.com/api/docs/guides/fast-mode",
         "https://developers.openai.com/api/docs/models/gpt-5.6-sol",
@@ -18,8 +19,9 @@ export const MODEL_COST_PRICING_SNAPSHOT = {
         "OpenAI Standard pricing; requests above 272,000 total input/cache tokens use 2x input/cache and 1.5x output pricing.",
         "Anthropic Standard pricing with the 5-minute cache-write rate; Batch, Flex, regional, tool, and subscription charges are excluded.",
         "variant and serviceTier are separate; model variants do not alter the selected Standard token price.",
+        "GPT-6.1 Terra Standard uses $1 input, $0.05 cache read, $1.25 30-minute cache write and $5 output per 1M tokens; subscription usage/limits are separate from this API-equivalent estimate.",
         "GPT-6 Sol, GPT-6.1 Sol, and GPT-6 Luna use their official Standard schedules; the OpenCode gpt-6-luna-fast selector sends API model gpt-6-luna with priority service tier.",
-        "For GPT-6 Sol, GPT-6.1 Sol, and GPT-6 Luna, Fast mode (service_tier fast or priority, including the gpt-6.1-sol-fast selector) multiplies applicable Standard rates by 2, including cache and long-context multipliers.",
+        "For GPT-6 Sol, GPT-6.1 Sol, GPT-6.1 Terra, and GPT-6 Luna, Fast mode (service_tier fast or priority, including the gpt-6.1-sol-fast selector) multiplies applicable Standard rates by 2, including cache and long-context multipliers.",
         "gpt-5.6-luna-fast is retained only as a compatibility price for historical host catalog entries; it has no published model page and is exactly twice the GPT-5.6 Luna Standard schedule.",
     ],
 };
@@ -27,6 +29,7 @@ const OPENAI = {
     "gpt-6-astra": { input: 10, cacheRead: 1, cacheWrite: 12.5, output: 50 },
     "gpt-6-sol": { input: 2, cacheRead: 0.2, cacheWrite: 2.5, output: 10 },
     "gpt-6.1-sol": { input: 2, cacheRead: 0.1, cacheWrite: 2.5, output: 10 },
+    "gpt-6.1-terra": { input: 1, cacheRead: 0.05, cacheWrite: 1.25, output: 5 },
     "gpt-6-luna": { input: 0.1, cacheRead: 0.01, cacheWrite: 0.125, output: 0.5 },
     "gpt-5.6-sol": { input: 4, cacheRead: 0.4, cacheWrite: 5, output: 20 },
     "gpt-5.6": { input: 4, cacheRead: 0.4, cacheWrite: 5, output: 20 },
@@ -49,7 +52,7 @@ export function estimateModelUsageCost(usage) {
     const sol61FastAlias = usage.providerID?.toLowerCase() === "openai" && usage.modelID === "gpt-6.1-sol-fast";
     const alias = lunaFastAlias || sol61FastAlias;
     const fast = alias || (["fast", "priority"].includes(usage.serviceTier?.toLowerCase() ?? "") &&
-        usage.providerID?.toLowerCase() === "openai" && ["gpt-6-sol", "gpt-6.1-sol", "gpt-6-luna"].includes(usage.modelID ?? ""));
+        usage.providerID?.toLowerCase() === "openai" && ["gpt-6-sol", "gpt-6.1-sol", "gpt-6.1-terra", "gpt-6-luna"].includes(usage.modelID ?? ""));
     if (usage.serviceTier !== undefined && !(fast ? ["standard", "default", "fast", "priority"] : ["standard", "default"]).includes(usage.serviceTier.toLowerCase())) {
         return { status: "unpriced", reason: "unsupported-service-tier" };
     }

package/dist/plugin/profiled.js

@@ -36,7 +36,8 @@ const SERIAL_OPTIONAL_ARGUMENTS = new Map([
             "scope", "trigger", "cause", "prevention", "evidence", "evidenceRef", "id", "promotedRef", "confirmation",
         ])],
 ]);
-const PREVIEW_WORKER_ROUTE = Object.freeze({ model: "openai/gpt-6-luna-fast", variant: "max" });
+const PREVIEW_WORKER_ROUTE = Object.freeze({ model: "openai/gpt-6.1-terra", variant: "max" });
+const PREVIEW_LUNA_WORKER_ROUTE = Object.freeze({ model: "openai/gpt-6-luna-fast", variant: "max" });
 const PREVIEW_SCOUT_ROUTE = Object.freeze({ model: "openai/gpt-6-luna-fast", variant: "max" });
 /**
  * Contract authorship, not throughput. Qualification observed a cheaper operations model emit
@@ -49,7 +50,7 @@ const PREVIEW_PRIMARY_ROUTE = Object.freeze({ model: "openai/gpt-6.1-sol", varia
 const PREVIEW_REVIEW_ROUTE = Object.freeze({ model: "openai/gpt-6.1-sol", variant: "xhigh" });
 /** Every preview route the profile can bind a role to. Catalog declaration reads this one list. */
 const PREVIEW_ROUTES = Object.freeze([
-    PREVIEW_PRIMARY_ROUTE, PREVIEW_WORKER_ROUTE, PREVIEW_SCOUT_ROUTE, PREVIEW_OPERATIONS_ROUTE,
+    PREVIEW_PRIMARY_ROUTE, PREVIEW_WORKER_ROUTE, PREVIEW_LUNA_WORKER_ROUTE, PREVIEW_SCOUT_ROUTE, PREVIEW_OPERATIONS_ROUTE,
     PREVIEW_REVIEW_ROUTE,
 ]);
 function missionReportReview(mission, runID) {
@@ -1492,8 +1493,8 @@ export function createProfiledPlugin(profile, assetVersion) {
         }
         async function stop(root, reason, retireRoot = true) {
             const directActor = (await operators.read(root))?.units.find(unit => unit.status === "running" && unit.directExecution)?.directExecution?.actor;
-            if (directActor)
-                await activateDirect(directActor);
+            // Stopping owned execution does not write its artifacts. Its existing reservation
+            // is reconciled by cancelChildren, even when activation never acquired a writer.
             if (retireRoot)
                 retired.add(root);
             if (retireRoot)
@@ -2289,7 +2290,9 @@ export function createProfiledPlugin(profile, assetVersion) {
                     record(part.state.metadata) && ["running", "pending", "unknown"].includes(String(part.state.metadata.status)))) {
                     return JSON.stringify({ status: "direct-handoff-awaits-tools", next_action: "Finish the existing native command before handing off this unit; retain its result and continue here or delegate afterwards." });
                 }
-                await activateDirect(actor);
+                // Yielding execution is management, not an artifact write. In particular, a
+                // direct admission can exist even when its writer could not bind because another
+                // live owner overlaps. Do not require that contested binding to hand it off.
                 const mission = await missions.retainContext(root, actor, history, notes);
                 await operators.handoffDirect(root, actor, await missions.handoffContext(mission));
             }
@@ -3599,7 +3602,7 @@ export function createProfiledPlugin(profile, assetVersion) {
                     else if (args.status === "blocked" || args.status === "needs-decision") {
                         // This owner is stopping its direct work, not validating it. Keep the
                         // failed checks/candidate and settle its existing reservation once.
-                        const direct = await activateDirect(context.sessionID);
+                        const direct = await directExecution(context.sessionID);
                         if (direct)
                             await control.failDirectUnit(root, context.sessionID);
                     }
@@ -3811,7 +3814,7 @@ export function createProfiledPlugin(profile, assetVersion) {
                     const agent = record(agents[name]) ? agents[name] : undefined;
                     if (!agent)
                         continue;
-                    const route = role === "dog-scout" ? PREVIEW_SCOUT_ROUTE : PREVIEW_WORKER_ROUTE;
+                    const route = role === "dog-scout" ? PREVIEW_SCOUT_ROUTE : PREVIEW_LUNA_WORKER_ROUTE;
                     agent.model ??= route.model;
                     if (agent.variant === undefined && agent.model === route.model)
                         agent.variant = route.variant;
@@ -4014,7 +4017,11 @@ export function createProfiledPlugin(profile, assetVersion) {
                     throw new Error(RUNTIME_PROFILE_SESSION_INACTIVE);
                 if (request.tool.startsWith("sortie_") && !request.tool.startsWith(profile.toolPrefix))
                     throw new Error("runtime-profile-tool-mismatch");
-                const direct = await activateDirect(request.sessionID);
+                // Management can reconcile, yield, stop or report the owner's admission.
+                // They must remain reachable when its artifact binding is the thing in conflict.
+                // Ordinary tools and the parts of planning that actually write still activate it.
+                const direct = request.tool === status || request.tool === planUnits || request.tool === cancel || request.tool === submitMission
+                    ? await directExecution(request.sessionID) : await activateDirect(request.sessionID);
                 const writer = who.role === "dog-worker" || !!direct || !!await reviewerCorrection(request.sessionID);
                 if (who.role === "dog-reviewer" && writer) {
                     const correction = (await reviewerCorrection(request.sessionID));

package/dist/plugin/v2.js

@@ -90,6 +90,22 @@ function legacyPart(value, session, message) {
             ...(time ? { time: { start: time.ran ?? time.created, end: time.completed } } : {}),
             ...(output === undefined ? {} : { output }) } };
 }
+/** Resolved V2 files carry the original bytes, even when their source URI is no longer readable. */
+function legacyUserParts(text, files, session, message) {
+    return [{ id: `${message}-text`, sessionID: session, messageID: message, type: "text", text },
+        ...array(files).flatMap((file, index) => {
+            if (!record(file))
+                return [];
+            const mime = string(file.mime) ?? "application/octet-stream";
+            const url = typeof file.data === "string" ? `data:${mime};base64,${file.data}`
+                : string(file.uri); // The prompt hook runs before URI resolution.
+            if (!url)
+                return [];
+            return [{ id: `${message}-file-${index}`, sessionID: session, messageID: message, type: "file", mime, url,
+                    ...(typeof file.name === "string" ? { filename: file.name } : {}),
+                    ...(record(file.source) ? { source: file.source } : {}) }];
+        })];
+}
 function legacyMessage(value, session, agent) {
     if (!record(value) || typeof value.id !== "string" || typeof value.type !== "string")
         return undefined;
@@ -97,7 +113,7 @@ function legacyMessage(value, session, agent) {
     const originalAgent = record(value.metadata) ? string(value.metadata.agent) ?? agent : agent;
     if (value.type === "user") {
         return { info: { id, sessionID: session, role: "user", agent: originalAgent, time: value.time },
-            parts: [{ id: `${id}-text`, sessionID: session, messageID: id, type: "text", text: String(value.text ?? "") }] };
+            parts: legacyUserParts(String(value.text ?? ""), value.files, session, id) };
     }
     if (value.type !== "assistant") {
         if (value.type === "synthetic")
@@ -289,8 +305,22 @@ function toolContentText(content) {
     return values.length === 0 ? undefined : values.join("\n");
 }
 function replaceToolContent(result, text) {
-    return typeof result.content === "string" ? { ...result, content: text }
-        : { ...result, content: [{ type: "text", text }] };
+    if (toolContentText(result.content) === text)
+        return result;
+    if (typeof result.content === "string")
+        return { ...result, content: text };
+    let replaced = false;
+    const content = array(result.content).flatMap(item => {
+        if (!record(item) || item.type !== "text" || typeof item.text !== "string")
+            return [item];
+        if (replaced)
+            return [];
+        replaced = true;
+        return [{ ...item, text }];
+    });
+    // Only the legacy textual projection changed. Native file/image blocks must still reach
+    // the host's model-input lowering, including images returned through Code Mode.
+    return { ...result, content: replaced ? content : [...content, { type: "text", text }] };
 }
 function assistantText(message) {
     if (!record(message) || message.type !== "assistant" || message.finish !== "stop" || typeof message.id !== "string")
@@ -542,7 +572,7 @@ async function registerV2Hooks(context, hooks, background, liveState) {
             ["dog-reviewer-v010", "gpt-6.1-sol", "xhigh"],
             ["dog-scout-v010", "gpt-6-luna-fast", "max"],
             ["dog-luna-worker-v010", "gpt-6-luna-fast", "max"],
-            ["dog-worker-v010", "gpt-6-luna-fast", "max"],
+            ["dog-worker-v010", "gpt-6.1-terra", "max"],
         ])
             editor.update(name, agent => { agent.model ??= { providerID: "openai", id, variant }; });
         editor.update("dog-reviewer-v010", agent => {
@@ -784,7 +814,7 @@ async function registerV2Hooks(context, hooks, background, liveState) {
                 selectedChildModels.add(String(event.sessionID));
             }
             const output = { message: { id: String(event.messageID ?? ""), agent: string(info.agent), model },
-                parts: [{ type: "text", text: legacyText }] };
+                parts: legacyUserParts(legacyText, event.prompt.files, String(event.sessionID), String(event.messageID ?? "")) };
             await hooks["chat.message"]({ sessionID: String(event.sessionID), messageID: String(event.messageID ?? ""),
                 ...(typeof info.agent === "string" ? { agent: info.agent } : {}), model }, output);
             await reconcileCorrectionPermissions(String(event.sessionID));

package/dist/runtime-assets-v010.js

@@ -427,7 +427,8 @@ export const runtimeAssets = Object.freeze([
             content = content.replace("mode: subagent\n", "mode: subagent\nhidden: true\n");
         }
         if (["dog-worker", "dog-luna-worker", "dog-scout"].includes(asset.name)) {
-            content = content.replace("mode: subagent\n", "mode: subagent\nmodel: openai/gpt-6-luna-fast#max\n");
+            const model = asset.name === "dog-worker" ? "openai/gpt-6.1-terra#max" : "openai/gpt-6-luna-fast#max";
+            content = content.replace("mode: subagent\n", `mode: subagent\nmodel: ${model}\n`);
         }
         if (["dog-reviewer", "dog-advisor"].includes(asset.name)) {
             content = content.replace("mode: subagent\n", "mode: subagent\nmodel: openai/gpt-6.1-sol#xhigh\n");

package/dist/runtime-mission-assets.js

@@ -155,6 +155,17 @@ ${MISSION_DEPENDENCY_WORKFLOW}
     SAME mission before any Worker starts, retains its Coordinator and cumulative spend, and checks
     old children before preparing a Worker. This is not permission to discard unchanged acceptance:
     if the scope is uncertain, ask the user which requirements remain instead of inferring a replacement.
+    When execution management is inconsistent or manifest-overlap blocks the owning direct writer,
+    explain the observed stop, what can be reconciled, and that the existing requirements, results,
+    checks and cumulative spend remain. Offer to recover and continue in ordinary conversation;
+    an ordinary affirmative reply or a request to self-recover is enough. No hidden command or ID is
+    needed from the user, and ordinary autonomous recovery does not require a new approval.
+    Inspect operator_status to reconcile only native-proven ended work. If this session owns a quiescent
+    direct admission that cannot bind, plan_units with executor: "worker" and no units yields that same
+    unit and reservation without acquiring the contested writer. Explain the result and continue the
+    returned Task when its write scope is available. A genuinely live or unproven other owner is not
+    released: wait or separate work, and do not dispatch repeated conflicting attempts. Do not invent
+    terminal state, validation success, a fresh Mission, or a budget reset.
     If a Coordinator reports mission-replan-worker-still-active after an interrupted native Task,
     it cannot call cancel_operator (root-only). Inspect operator_status and the native Task outcome.
     If the Worker is still active, wait, not duplicate. If the user chose to stop/replan, root calls
@@ -216,7 +227,17 @@ ${MISSION_DEPENDENCY_WORKFLOW}
    and cumulative budget. Use its Coordinator Task only for real coordination, contract discovery or
    correction that cannot be handled directly. Never replace an active Worker or ask for routine approval.
 4. After the review decision (including a justified low-risk skip), compare the completion candidate
-    against the original request, real source and observed evidence before final acceptance.
+     against the original request, real source and observed evidence before final acceptance.
+     Include relevant original attachments: continuity sources retain ordered, message-linked read references
+     to the original bytes, not a descriptive substitute. Retrieve/inspect those references when needed.
+     Judge what each existing check actually exercises. Functional PASS, Review PASS, or honestly retaining
+     an unverified caveat does not establish a required visual comparison. If a mandatory outcome is unmet,
+     continue authorized retrieval, inspection or correction without waiting for a user reminder or approval.
+     Choose the effective method/author; reuse applicable evidence rather than repeat every check yourself.
+     Optional caveats and valid alternative verification can permit acceptance: the word unverified alone
+     is not a stop rule. This adds no approval gate, proof file, extra Reviewer, universal screenshot duty
+     or pixel-exact matching requirement. Keep autonomy, efficiency and visibility: report the actual gap
+     and next action, not an unsupported completion claim or a new restriction on ordinary work.
     For a Coordinator ready candidate, use operator_status's acceptance_summary: verbatim original
     requests, anchored cumulative formal validation, current review disposition (author self-recheck
     is not independent approval) and recorded delivery state.
@@ -484,6 +505,10 @@ in this SAME session when your existing context makes that efficient. You are no
 Use the Operator's preferred_executor as a starting recommendation. Start useful work with available context;
 do not investigate merely to confirm the recommendation. Adapt through actual work, honoring explicit user choices.
 Use plan_units with executor: "self" for direct work, or start_direct_unit for an already prepared unit.
+If a direct admission cannot bind because of manifest-overlap, operator_status and plan_units with
+executor: "worker" and no units remain management operations: yield only this quiescent owner, retain
+its checks and reservation, and wait for a real conflicting writer rather than evicting it. Continue
+the same Mission and actual remaining work; do not create another Mission to clear the conflict.
 The host binds your native check results to the candidate; finish_direct_unit closes that work without
 a Worker handoff. Independent Review and Operator final acceptance still follow. This is not limited to
 micro-edits. Delegate to the configured Worker when that reduces work/cost or provides useful separation.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sortie-dogs",
-  "version": "0.14.5",
+  "version": "0.14.7",
   "description": "Bounded agent harness and validated orchestration loop plugin for OpenCode",
   "keywords": [
     "opencode",