@gajae-code/coding-agent 0.4.4 → 0.5.0

package/CHANGELOG.md

@@ -2,6 +2,89 @@
 
 ## [Unreleased]
 
+## [0.5.0] - 2026-06-13
+
+### Fixed
+
+- Fixed forced `tool_choice` 400s ("tool_choice forces tool use is not compatible with this model") looping after `ast_edit` previews: named queue directives (resolve protocol, eager `todo_write` enforcement, subagent `yield` reminders) now enqueue only when the model supports exact named forcing; otherwise they degrade silently to the existing steer reminder without a forced `tool_choice`, and a runtime-discovered incapability drops the in-flight directive instead of requeueing it.
+- `models.yml` compat blocks now accept the `toolChoiceSupport` enum (`none`/`auto`/`required`/`named`) alongside the legacy `supportsToolChoice`/`supportsForcedToolChoice` booleans, mirrored in the generated JSON schema.
+### Added
+
+- Made `/model` open to a preset-first landing view: provider-grouped presets with live auth checkmarks, highlight-to-expand tiers, a full clamped role→model preview before applying, and a session/default apply scope choice; typing still jumps straight to model search, "Browse all models" opens the classic tabbed selector, and temporary-only quick-switch bypasses the landing entirely.
+- Rebuilt the builtin model profile catalog as 25 profiles: `codex-{eco,medium,pro}` on `gpt-5.5` effort spreads, a single `opencodego` preset, `claude-opus`, `{glm,kimi-coding-plan,mimo,grok,cursor,minimax}-{eco,medium,pro}` trios with thinking levels clamped to provider support, and `opus-codex`/`codex-opencodego` combos. Legacy profile names (including the `*-standard` family and retired Fable presets) were removed clean-break and now fail with the available-profile listing.
+- Added a post-`/login` smart preset recommendation: when login succeeds and no profile is active, prompts "Apply <preset> now?" (session-only on confirm); when a profile is active, prints a one-line hint instead. The active profile is tracked in-memory on the session with rollback-safe activation.
+- Bundled `kimi-code/kimi-k2.7-code` and `minimax-code/minimax-v3` model entries; MiniMax presets use the canonical `minimax-code` provider id throughout.
+- Added a harness receipt JSONL spool exporter for gajae receipt-runtime interop: configured `gjc harness --receipt-spool-dir <dir>` / `GJC_RECEIPT_SPOOL_DIR` now appends persisted native `ReceiptEnvelope` records as `{cursor,envelope}` lines to `spool.jsonl`, with restart-safe 12-digit cursors and installed-package smoke coverage (#545).
+- Added Gajae Trinity compatibility golden fixtures and tests that pin ReceiptEnvelope hash basis, validator compatibility, and replayable RPC exchange shape for downstream receipt-runtime interop.
+- Optimization Suite v3 Lane 1 (RSS): large resident text in persisted sessions is now backed by an ephemeral session-scoped disk cache (`EphemeralBlobStore`) instead of being pinned in JS heap for the whole session lifetime; canonical JSONL persistence, reload, and export semantics are byte-identical (resident refs never persist). Missing resident text cache blobs now surface a typed `ResidentBlobMissingError` instead of silently leaking `blob:sha256:` refs into provider payloads, UI, or exports. `getEntries()`/`buildSessionContext()` are served from revision-keyed WeakRef caches below the public ownership boundary (callers still receive caller-owned copies). Fixture retained heap −82%, RSS −55%, warm `getEntries()` p95 −80% on 10k-entry sessions; one-shot `exportFromFile()` now closes its session manager.
+- Added process-isolated deterministic TUI render-golden capture and fixtures for interactive editor overlays, rich-text resizing, multiplexer viewport repaint, sixel image line preservation, Termux height diffs, and transcript shrink/clear regressions.
+
+### Removed
+
+- Removed the hardcoded OpenAI Codex role-preset action from the model selector; builtin model profiles are now the only preset concept.
+- Removed retired Fable model profiles (`claude-fable`, `fable-codex`) after `claude-fable-5` was removed upstream.
+
+### Changed
+
+- Optimization Suite v3 Lane 3 (serialization): session-switch message comparison now uses per-message cached source strings + xxHash64 as an accelerator (source-string compare remains the authority; collision fallback tested) — unchanged-session compares −95% median. The secret obfuscator precomputes a longest-first combined regex (single-pass replace, −70% median/−77% p95 on 100 secrets × 1MiB) with a conservative sequential fallback whenever secrets overlap each other or any replacement/placeholder contains a secret — output bytes are identical in all cases. Intra-line diff rendering gains byte-identical fast paths for identical lines and whitespace-token-aligned prefix/suffix spans (identical −67%, single-token −60%; long lines skip the scan). Mental-model LCS keeps legacy dense-DP tie-break semantics (a Hunt-Szymanski variant was rejected for changing rendered bytes). Provider-visible fork-context seeds use JSON-semantic cloning instead of structuredClone.
+- Tightened tool-block rendering to remove vertical padding and rely on Spacer-only separation, reducing transcript noise while preserving stable render-golden output.
+- Improved the Bun runtime version guard diagnostic: when the Bun running `gjc` is older than the required version, the error now names the exact detected Bun runtime path and prints a platform-specific upgrade and PATH fix (Windows gets the `irm bun.sh/install.ps1|iex` reinstall plus a `%USERPROFILE%\.bun\bin` PATH hint) instead of a bare `bun upgrade` (#525).
+- Aligned the `codex-standard` and `codex-pro` model profiles on the `openai-codex/gpt-5.5` baseline so they no longer default to stale mixed model generations (`gpt-5.4`, `gpt-5.2-codex`, `gpt-5.1-codex-max`, `gpt-5.3-codex-spark`); the profiles now differentiate purely by per-role reasoning effort (#532).
+- Reduced the default RPC `get_state` payload by omitting static `dumpTools` and `systemPrompt` fields unless requested via `include: ["tools", "systemPrompt"]` (#539).
+- Updated `/model` documentation and generated docs index for the rebuilt preset catalog and preset-first selector.
+
+### Fixed
+
+- Tightened the Windows/psmux tmux provider boundary: `gjc team` now honors `GJC_TMUX_COMMAND` (not just `GJC_TEAM_TMUX_COMMAND`) so the team leader resolves the same multiplexer as `gjc session`/`gjc --tmux`; and when a multiplexer lists a session that lacks GJC's `@gjc-profile` ownership tag, `gjc session status` now returns `gjc_tmux_session_untagged` with a `detail` hint and `gjc team` reports the same cause, instead of a bare `gjc_tmux_session_not_found` / `unmanaged_tmux_session`. Documented that alternative multiplexers such as psmux on Windows are not fully supported because they do not round-trip tmux user options (#531).
+- Hardened RPC stdio lifecycle behavior: `gjc --mode rpc` now reports malformed JSONL frames as parse-error responses without killing the session, flushes durable session state before exiting on EOF/shutdown, and has red-team coverage for attached persistence, reload, malformed-frame recovery, and concurrent child-session isolation.
+- Hardened the harness RPC submit/router contract so `submit` is no longer advertised or accepted during finalizing/non-idle lifecycle windows, non-idle RPC state reports `submitted:false` with a retryable gate, and degraded owner endpoints fall back to `owner-not-live` without false acceptance (#544).
+- Ran estimated context maintenance before sending a new prompt, including tool-output pruning and threshold compaction, so large tool results appended after the last assistant turn cannot push the next model request over the context window.
+- `gjc team` now self-heals a missing `@gjc-profile` ownership tag when the current leader pane was genuinely launched by `gjc --tmux` (detected via `GJC_TMUX_LAUNCHED=1`): the session is re-tagged with `set-option` and startup proceeds, instead of hard-failing with `unmanaged_tmux_session` after a mid-startup attach failure or registry race stripped the tag. Sessions without the GJC launch marker are still rejected unchanged, so foreign tmux sessions cannot be hijacked.
+- Subagent task receipts and live render output now warn when requested role-agent models are substituted by auth fallback or provider-reported assistant model mismatch, including session model-change annotations for server-side substitutions (#559).
+- Converted Cursor wire shell timeouts from millisecond values to bash-tool seconds so delegated Cursor-native shell calls honor the expected timeout units.
+- Fixed pi-shell bash fixups on multibyte UTF-8 commands by converting parser source indexes to byte offsets before stripping `head`/`tail` pipelines.
+
+## [0.4.5] - 2026-06-12
+
+### Added
+
+- Added a dim `(ctrl+s to observe sessions)` discoverability hint under the `subagent` await panel header while any awaited subagent is still running, pointing to the full session observer overlay; the hint shows in both collapsed and expanded states and disappears once no subagent is running.
+- Added a `phase-rollup` receipt family (receipt-of-receipts) to the harness control plane: a hash-sealed rollup that supersedes N child task receipts at a lifecycle boundary, preserving per-child `{id, status, outputUri, outputSha256, receiptSha256, tokens, costTotal, clonedTokens, lowRoi}` pointers plus aggregate ROI totals, with a pure deterministic builder (`buildPhaseRollupReceipt`) and fail-closed semantic validation.
+- Added a pure receipt-ingestion fast path (`ingestReceipts`): fail-closed batch validation + lifecycle transition computation via the existing state machine, plus a deterministic model-facing digest hard-capped at 280 chars — groundwork for LLM-free receipt routing.
+- Added advisory spawn-ROI reconciliation (`reconcileSpawnRoi`) and deterministic fork-context mode advice (`adviseForkContextMode`) surfaced in task receipts without changing task success semantics.
+- Added the Grok Build provider contract design document.
+
+### Changed
+
+- Reduced compiled CLI startup and native bundle pressure with default-small grammar loading, tokenizer tiering, and compiled fast-help paths.
+- Preserved dev/main release metadata and changelog consistency for the 0.4.5 lockstep release.
+
+- Added native `gjc ultragoal steer --kind` support for documented steering mutations beyond `add_subgoal`, including split, reorder, wording revision, ledger annotation, and blocked-goal supersession contracts with structured audit expectations.
+### Fixed
+
+- Kept the unified `goal` tool registered and active by default whenever `goal.enabled` is true, including explicit tool subsets and `gjc ultragoal create-goals` arming flows.
+- Restored no-argument `gjc` interactive startup instead of launching help.
+- Rendered and executed Cursor-native tool calls, including detached/native handler paths and empty-pattern composer grep guards.
+- Tool-output pruning no longer rewrites already-sent provider-facing history mid prompt-cache epoch and now persists pruned message updates back into canonical session storage.
+- Preserved provider abort root causes in the final TUI abort label, kept replay rendering idempotent, and added a `PI_STREAM_IDLE_TIMEOUT_MS` remediation hint when stream idle watchdogs fire.
+- Hardened harness owner recovery/finalize paths and submit-prompt-file handling.
+
+## [0.4.4] - 2026-06-10
+
+### Added
+
+- Made coordinator turns event-backed, adding an MCP coordinator server and the `gjc coordinator` / `gjc mcp-serve` commands backed by durable turn/session state (#479).
+- Surfaced awaited sub-agent live streaming status in the await panel (#475).
+
+### Changed
+
+- Refreshed the README brand images (#477).
+
+### Fixed
+
+- Persisted ralplan role-agent artifacts via the CLI, returning receipt-only output to the caller (#474).
+- Collected the `ask` tool "Other" custom input inline below the option list (#476).
+
 ## [0.4.3] - 2026-06-10
 
 ### Added

package/dist/types/cli/fast-help.d.ts

@@ -0,0 +1 @@
+export declare function getExtraHelpText(): string;

package/dist/types/cli/setup-cli.d.ts

@@ -18,6 +18,8 @@ export interface SetupCommandArgs {
         repo?: string;
         profile?: string;
         sessionCommand?: string;
+        noWorktree?: boolean;
+        worktreeName?: string;
         stateRoot?: string;
         mutation?: string[];
         artifactByteCap?: string;

package/dist/types/commands/harness.d.ts

@@ -14,6 +14,9 @@ export default class Harness extends Command {
             description: string;
             default: string;
         };
+        "prompt-file": import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
+            description: string;
+        };
         session: import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
             char: string;
             description: string;
@@ -31,6 +34,9 @@ export default class Harness extends Command {
             description: string;
             default: boolean;
         };
+        "receipt-spool-dir": import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
+            description: string;
+        };
     };
     static examples: string[];
     run(): Promise<void>;

package/dist/types/commands/setup.d.ts

@@ -43,6 +43,12 @@ export default class Setup extends Command {
         "session-command": import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
             description: string;
         };
+        "no-worktree": import("@gajae-code/utils/cli").FlagDescriptor<"boolean"> & {
+            description: string;
+        };
+        "worktree-name": import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
+            description: string;
+        };
         "state-root": import("@gajae-code/utils/cli").FlagDescriptor<"string"> & {
             description: string;
         };

package/dist/types/config/model-profile-activation.d.ts

@@ -3,15 +3,22 @@ import type { Api, Model } from "@gajae-code/ai";
 import type { AgentSession } from "../session/agent-session";
 import { type ModelRegistry } from "./model-registry";
 import type { Settings } from "./settings";
+type ModelProfileActivationSession = Pick<AgentSession, "model" | "thinkingLevel" | "sessionId"> & {
+    setModelTemporary?: AgentSession["setModelTemporary"];
+    setActiveModelProfile?: (name: string | undefined) => void;
+    getActiveModelProfile?: () => string | undefined;
+};
 export interface PrepareModelProfileActivationOptions {
-    session: Pick<AgentSession, "model" | "thinkingLevel" | "sessionId">;
+    session: ModelProfileActivationSession;
     modelRegistry: Pick<ModelRegistry, "getModelProfile" | "getModelProfiles" | "getAvailableModelProfileNames" | "getApiKeyForProvider" | "getAll" | "resolveCanonicalModel" | "getCanonicalVariants" | "getCanonicalId">;
     settings: Pick<Settings, "get">;
     profileName: string;
 }
 export interface PreparedModelProfileActivation {
     profileName: string;
-    session: Pick<AgentSession, "model" | "thinkingLevel" | "sessionId" | "setModelTemporary">;
+    session: ModelProfileActivationSession & {
+        setModelTemporary: AgentSession["setModelTemporary"];
+    };
     settings: Pick<Settings, "get" | "override" | "set" | "flush">;
     previousModel: Model<Api> | undefined;
     previousThinkingLevel: ThinkingLevel | undefined;
@@ -19,6 +26,7 @@ export interface PreparedModelProfileActivation {
     defaultModel: Model<Api> | undefined;
     defaultThinkingLevel: ThinkingLevel | undefined;
     agentModelOverrides: Record<string, string>;
+    previousActiveModelProfile: string | undefined;
 }
 export declare function formatModelProfileCredentialError(profileName: string, providers: readonly string[]): string;
 export declare function prepareModelProfileActivation(options: PrepareModelProfileActivationOptions): Promise<PreparedModelProfileActivation>;
@@ -28,3 +36,4 @@ export declare function applyPreparedModelProfileActivation(prepared: PreparedMo
 export declare function activateModelProfile(options: PrepareModelProfileActivationOptions, applyOptions?: {
     persistDefault?: boolean;
 }): Promise<void>;
+export {};

package/dist/types/config/model-profiles.d.ts

@@ -14,6 +14,13 @@ export interface ResolvedProfileBinding {
 export declare function deriveModelProfileMappedProviders(definition: Pick<ModelProfileDefinition, "modelMapping">): string[];
 export declare function aggregateModelProfileRequiredProviders(requiredProviders: readonly string[], definition: Pick<ModelProfileDefinition, "modelMapping">): string[];
 export declare const BUILTIN_MODEL_PROFILES: readonly ModelProfileDefinition[];
+export interface ModelProfilePresentation {
+    displayName: string;
+    providerGroup: string;
+}
+export declare function getModelProfilePresentation(name: string): ModelProfilePresentation;
+export declare function groupModelProfilesForPresetLanding(profiles: ReadonlyMap<string, ModelProfileDefinition>): Map<string, ModelProfileDefinition[]>;
+export declare function recommendModelProfileForProvider(providerId: string, profiles: ReadonlyMap<string, ModelProfileDefinition>): ModelProfileDefinition | undefined;
 export declare function mergeModelProfiles(userProfiles?: ModelsConfig["profiles"]): Map<string, ModelProfileDefinition>;
 export declare function resolveProfileBindings(definition: ModelProfileDefinition): ResolvedProfileBinding;
 export declare function formatAvailableProfileNames(profiles: ReadonlyMap<string, ModelProfileDefinition>): string;

package/dist/types/config/model-registry.d.ts

@@ -70,6 +70,8 @@ export declare const ModelsConfigFile: ConfigFile<{
             allowsSyntheticReasoningContentForToolCalls?: boolean | undefined;
             requiresAssistantContentForToolCalls?: boolean | undefined;
             supportsToolChoice?: boolean | undefined;
+            supportsForcedToolChoice?: boolean | undefined;
+            toolChoiceSupport?: "auto" | "named" | "none" | "required" | undefined;
             disableReasoningOnForcedToolChoice?: boolean | undefined;
             disableReasoningOnToolChoice?: boolean | undefined;
             thinkingFormat?: "openai" | "openrouter" | "qwen" | "qwen-chat-template" | "zai" | undefined;
@@ -145,6 +147,8 @@ export declare const ModelsConfigFile: ConfigFile<{
                 allowsSyntheticReasoningContentForToolCalls?: boolean | undefined;
                 requiresAssistantContentForToolCalls?: boolean | undefined;
                 supportsToolChoice?: boolean | undefined;
+                supportsForcedToolChoice?: boolean | undefined;
+                toolChoiceSupport?: "auto" | "named" | "none" | "required" | undefined;
                 disableReasoningOnForcedToolChoice?: boolean | undefined;
                 disableReasoningOnToolChoice?: boolean | undefined;
                 thinkingFormat?: "openai" | "openrouter" | "qwen" | "qwen-chat-template" | "zai" | undefined;
@@ -216,6 +220,8 @@ export declare const ModelsConfigFile: ConfigFile<{
                 allowsSyntheticReasoningContentForToolCalls?: boolean | undefined;
                 requiresAssistantContentForToolCalls?: boolean | undefined;
                 supportsToolChoice?: boolean | undefined;
+                supportsForcedToolChoice?: boolean | undefined;
+                toolChoiceSupport?: "auto" | "named" | "none" | "required" | undefined;
                 disableReasoningOnForcedToolChoice?: boolean | undefined;
                 disableReasoningOnToolChoice?: boolean | undefined;
                 thinkingFormat?: "openai" | "openrouter" | "qwen" | "qwen-chat-template" | "zai" | undefined;

package/dist/types/config/model-resolver.d.ts

@@ -133,6 +133,8 @@ export declare function resolveModelOverrideWithAuthFallback(modelPatterns: stri
     thinkingLevel?: ThinkingLevel;
     explicitThinkingLevel: boolean;
     authFallbackUsed: boolean;
+    requestedModel?: Model<Api>;
+    fallbackReason?: "auth_unavailable";
 }>;
 /**
  * Resolve a list of role patterns to the first matching model.

package/dist/types/config/models-config-schema.d.ts

@@ -30,6 +30,13 @@ export declare const OpenAICompatSchema: z.ZodObject<{
     allowsSyntheticReasoningContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
     requiresAssistantContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
     supportsToolChoice: z.ZodOptional<z.ZodBoolean>;
+    supportsForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
+    toolChoiceSupport: z.ZodOptional<z.ZodEnum<{
+        auto: "auto";
+        named: "named";
+        none: "none";
+        required: "required";
+    }>>;
     disableReasoningOnForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
     disableReasoningOnToolChoice: z.ZodOptional<z.ZodBoolean>;
     thinkingFormat: z.ZodOptional<z.ZodEnum<{
@@ -182,6 +189,13 @@ export declare const ModelOverrideSchema: z.ZodObject<{
         allowsSyntheticReasoningContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
         requiresAssistantContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
         supportsToolChoice: z.ZodOptional<z.ZodBoolean>;
+        supportsForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
+        toolChoiceSupport: z.ZodOptional<z.ZodEnum<{
+            auto: "auto";
+            named: "named";
+            none: "none";
+            required: "required";
+        }>>;
         disableReasoningOnForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
         disableReasoningOnToolChoice: z.ZodOptional<z.ZodBoolean>;
         thinkingFormat: z.ZodOptional<z.ZodEnum<{
@@ -288,6 +302,13 @@ export declare const ModelsConfigSchema: z.ZodObject<{
             allowsSyntheticReasoningContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
             requiresAssistantContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
             supportsToolChoice: z.ZodOptional<z.ZodBoolean>;
+            supportsForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
+            toolChoiceSupport: z.ZodOptional<z.ZodEnum<{
+                auto: "auto";
+                named: "named";
+                none: "none";
+                required: "required";
+            }>>;
             disableReasoningOnForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
             disableReasoningOnToolChoice: z.ZodOptional<z.ZodBoolean>;
             thinkingFormat: z.ZodOptional<z.ZodEnum<{
@@ -442,6 +463,13 @@ export declare const ModelsConfigSchema: z.ZodObject<{
                 allowsSyntheticReasoningContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
                 requiresAssistantContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
                 supportsToolChoice: z.ZodOptional<z.ZodBoolean>;
+                supportsForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
+                toolChoiceSupport: z.ZodOptional<z.ZodEnum<{
+                    auto: "auto";
+                    named: "named";
+                    none: "none";
+                    required: "required";
+                }>>;
                 disableReasoningOnForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
                 disableReasoningOnToolChoice: z.ZodOptional<z.ZodBoolean>;
                 thinkingFormat: z.ZodOptional<z.ZodEnum<{
@@ -575,6 +603,13 @@ export declare const ModelsConfigSchema: z.ZodObject<{
                 allowsSyntheticReasoningContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
                 requiresAssistantContentForToolCalls: z.ZodOptional<z.ZodBoolean>;
                 supportsToolChoice: z.ZodOptional<z.ZodBoolean>;
+                supportsForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
+                toolChoiceSupport: z.ZodOptional<z.ZodEnum<{
+                    auto: "auto";
+                    named: "named";
+                    none: "none";
+                    required: "required";
+                }>>;
                 disableReasoningOnForcedToolChoice: z.ZodOptional<z.ZodBoolean>;
                 disableReasoningOnToolChoice: z.ZodOptional<z.ZodBoolean>;
                 thinkingFormat: z.ZodOptional<z.ZodEnum<{

package/dist/types/config/settings-schema.d.ts

@@ -1,3 +1,4 @@
+import type { Effort } from "@gajae-code/ai/model-thinking";
 import { type SkillDiscoverySettings } from "./skill-settings-defaults";
 /** Unified settings schema - single source of truth for all settings.
  *
@@ -697,13 +698,13 @@ export declare const SETTINGS_SCHEMA: {
     };
     readonly defaultThinkingLevel: {
         readonly type: "enum";
-        readonly values: readonly import("@gajae-code/ai").Effort[];
-        readonly default: "high";
+        readonly values: readonly Effort[];
+        readonly default: Effort;
         readonly ui: {
             readonly tab: "model";
             readonly label: "Thinking Level";
             readonly description: "Reasoning depth for thinking-capable models";
-            readonly options: readonly import("../thinking").ThinkingLevelMetadata[];
+            readonly options: readonly import("../thinking-metadata").ThinkingLevelMetadata[];
         };
     };
     readonly hideThinkingBlock: {

package/dist/types/coordinator/contract.d.ts

@@ -1,4 +1,4 @@
 export declare const COORDINATOR_MCP_PROTOCOL_VERSION = "2024-11-05";
 export declare const COORDINATOR_MCP_SERVER_NAME = "gjc-coordinator-mcp";
-export declare const COORDINATOR_MCP_TOOL_NAMES: readonly ["gjc_coordinator_list_sessions", "gjc_coordinator_read_status", "gjc_coordinator_read_tail", "gjc_coordinator_list_questions", "gjc_coordinator_list_artifacts", "gjc_coordinator_read_artifact", "gjc_coordinator_read_coordination_status", "gjc_coordinator_start_session", "gjc_coordinator_send_prompt", "gjc_coordinator_submit_question_answer", "gjc_coordinator_read_turn", "gjc_coordinator_await_turn", "gjc_coordinator_report_status"];
+export declare const COORDINATOR_MCP_TOOL_NAMES: readonly ["gjc_coordinator_list_sessions", "gjc_coordinator_read_status", "gjc_coordinator_read_tail", "gjc_coordinator_list_questions", "gjc_coordinator_list_artifacts", "gjc_coordinator_read_artifact", "gjc_coordinator_read_coordination_status", "gjc_coordinator_register_session", "gjc_coordinator_start_session", "gjc_coordinator_send_prompt", "gjc_coordinator_submit_question_answer", "gjc_coordinator_read_turn", "gjc_coordinator_await_turn", "gjc_coordinator_report_status"];
 export type CoordinatorToolName = (typeof COORDINATOR_MCP_TOOL_NAMES)[number];

package/dist/types/coordinator-mcp/server.d.ts

@@ -8,10 +8,11 @@ interface JsonRpcRequest {
     method: string;
     params?: unknown;
 }
+type JsonRpcResult = any;
 interface JsonRpcResponse {
     jsonrpc: "2.0";
     id: string | number | null;
-    result?: any;
+    result?: JsonRpcResult;
     error?: {
         code: number;
         message: string;
@@ -30,6 +31,11 @@ interface SessionStartInput {
 interface CoordinatorServices {
     listSessions?: () => unknown[] | Promise<unknown[]>;
     startSession?: (input: SessionStartInput) => unknown | Promise<unknown>;
+    commandRunner?: (command: string[]) => Promise<{
+        exitCode: number;
+        stdout: string;
+        stderr: string;
+    }>;
 }
 interface CoordinatorMcpServerOptions {
     env?: NodeJS.ProcessEnv;
@@ -48,5 +54,5 @@ export declare function createCoordinatorMcpServer(options?: CoordinatorMcpServe
     handleJsonRpc: (request: JsonRpcRequest) => Promise<JsonRpcResponse>;
     handle: (request: JsonRpcRequest) => Promise<JsonRpcResponse>;
 };
-export declare function handleCoordinatorMcpRequest(request: JsonRpcRequest, options?: LegacyHandlerOptions): Promise<any>;
+export declare function handleCoordinatorMcpRequest(request: JsonRpcRequest, options?: LegacyHandlerOptions): Promise<JsonRpcResponse>;
 export declare function runCoordinatorMcpStdio(options?: CoordinatorMcpServerOptions): Promise<void>;

package/dist/types/gjc-runtime/team-runtime.d.ts

@@ -275,7 +275,6 @@ export declare function resolveGjcTeamStateRoot(cwd?: string, env?: NodeJS.Proce
 export declare function persistGjcTeamModeStateSummary(snapshot: GjcTeamSnapshot, cwd?: string): Promise<void>;
 export declare function recoverGjcTeamStaleClaims(teamName: string, cwd?: string, env?: NodeJS.ProcessEnv): Promise<GjcTeamLivenessRecoveryResult>;
 type GjcTeamTaskMetadataInput = Partial<Pick<GjcTeamTask, "owner" | "lane" | "required_role" | "allowed_roles" | "depends_on" | "blocked_by">>;
-export declare function resolveGjcTmuxCommand(env?: NodeJS.ProcessEnv): string;
 export declare function resolveGjcWorkerCommand(cwd?: string, env?: NodeJS.ProcessEnv): string;
 export type GjcWorkerCheckpointClassification = {
     kind: "clean";

package/dist/types/gjc-runtime/tmux-common.d.ts

@@ -21,6 +21,9 @@ export interface TmuxCommandResult {
 export type TmuxCommandRunner = (args: string[]) => TmuxCommandResult;
 export declare function envDisabled(value: string | undefined): boolean;
 export declare function resolveGjcTmuxCommand(env?: NodeJS.ProcessEnv): string;
+export declare const GJC_TMUX_UNTAGGED_REASON = "gjc_tmux_session_untagged";
+export declare function buildGjcTmuxUntaggedSessionHint(tmuxCommand: string): string;
+export declare function buildGjcTmuxUntaggedSessionError(sessionName: string, tmuxCommand: string): string;
 export declare function sanitizeTmuxToken(value: string): string;
 export declare function buildGjcTmuxSessionSlug(value: string): string;
 export declare function buildGjcTmuxSessionName(env?: NodeJS.ProcessEnv, context?: {

package/dist/types/harness-control-plane/finalize.d.ts

@@ -30,6 +30,11 @@ export interface FinalizeOptions {
     reviewOnly?: boolean;
     /** Operator/loop-supplied terminal review verdict (closed vocabulary). */
     verdict?: string | null;
+    /**
+     * Final assistant text from the live RPC owner, used to extract a closed-vocabulary verdict
+     * for review-only sessions when no explicit {@link verdict} is supplied. Never persisted raw.
+     */
+    assistantText?: string | null;
     /** Bounded PR/issue reference for the review target (e.g. "PR-414"). Never resolved from the live repo. */
     prTarget?: string | null;
     validationCommands?: ValidationCommandSpec[];

package/dist/types/harness-control-plane/owner.d.ts

@@ -11,7 +11,7 @@
  * Stateless `gjc harness` CLI calls reach the owner via {@link resolveOwner} + the endpoint.
  */
 import { type FinalizeChecks, type ValidationCommandSpec } from "./finalize";
-import type { HarnessRpc } from "./rpc-adapter";
+import { type HarnessRpc } from "./rpc-adapter";
 import { type SessionLease } from "./session-lease";
 export interface OwnerOptions {
     root: string;

package/dist/types/harness-control-plane/phase-rollup.d.ts

@@ -0,0 +1,23 @@
+/**
+ * Phase-boundary receipt rollup builder (receipt-of-receipts).
+ *
+ * At a harness lifecycle boundary, N child task receipts can be superseded by a
+ * single `phase-rollup` receipt that preserves per-child pointers (id, status,
+ * outputRef, sha256) plus aggregate ROI totals. The rollup is hash-sealed via
+ * the standard receipt envelope and validated fail-closed like every other
+ * family (see `validatePhaseRollup` in receipts.ts). Pure builder — no runtime
+ * injection behavior is changed here.
+ */
+import type { TaskResultReceipt } from "../task/receipt";
+import { type BuildReceiptInput, type PhaseRollupEvidence, type ReceiptEnvelope } from "./receipts";
+export interface BuildPhaseRollupInput {
+    receiptId: string;
+    sessionId: string;
+    source: string;
+    subject: BuildReceiptInput<PhaseRollupEvidence>["subject"];
+    phase: string;
+    children: readonly TaskResultReceipt[];
+    /** Supply for deterministic output; defaults to now. */
+    createdAt?: string;
+}
+export declare function buildPhaseRollupReceipt(input: BuildPhaseRollupInput): ReceiptEnvelope<PhaseRollupEvidence>;

package/dist/types/harness-control-plane/receipt-ingest.d.ts

@@ -0,0 +1,19 @@
+import type { ReceiptEnvelope } from "./receipts";
+import type { HarnessLifecycle, ReceiptFamily, SessionState } from "./types";
+export declare const RECEIPT_DIGEST_MAX_CHARS = 280;
+export declare const RECEIPT_FAMILY_LIFECYCLE_TARGETS: Partial<Record<ReceiptFamily, HarnessLifecycle>>;
+export interface ReceiptIngestResult {
+    accepted: ReceiptEnvelope<unknown>[];
+    rejected: {
+        receipt: ReceiptEnvelope<unknown>;
+        reasons: string[];
+    }[];
+    transitions: {
+        from: HarnessLifecycle;
+        to: HarnessLifecycle;
+        receiptId: string;
+    }[];
+    finalLifecycle: HarnessLifecycle;
+    digest: string;
+}
+export declare function ingestReceipts(state: SessionState, receipts: readonly ReceiptEnvelope<unknown>[]): ReceiptIngestResult;

package/dist/types/harness-control-plane/receipt-spool.d.ts

@@ -0,0 +1,19 @@
+import type { ReceiptEnvelope } from "./receipts";
+export declare const RECEIPT_SPOOL_DIR_ENV = "GJC_RECEIPT_SPOOL_DIR";
+export declare const RECEIPT_SPOOL_FILENAME = "spool.jsonl";
+export declare const RECEIPT_SPOOL_CURSOR_WIDTH = 12;
+export interface ReceiptSpoolRecord {
+    cursor: string;
+    envelope: ReceiptEnvelope<unknown>;
+}
+export interface ReceiptSpoolAppendResult {
+    cursor: string;
+    path: string;
+}
+export declare function withReceiptSpoolDir<T>(spoolDir: string, fn: () => Promise<T>): Promise<T>;
+export declare function resolveReceiptSpoolDir(env?: NodeJS.ProcessEnv): string | undefined;
+export declare function receiptSpoolPath(spoolDir: string): string;
+export declare function formatReceiptSpoolCursor(cursor: bigint): string;
+export declare function readHighestReceiptSpoolCursor(spoolDir: string): Promise<bigint>;
+export declare function appendReceiptToSpool(spoolDir: string, envelope: ReceiptEnvelope<unknown>): Promise<ReceiptSpoolAppendResult>;
+export declare function appendReceiptToConfiguredSpool(envelope: ReceiptEnvelope<unknown>, env?: NodeJS.ProcessEnv): Promise<ReceiptSpoolAppendResult | undefined>;

package/dist/types/harness-control-plane/receipts.d.ts

@@ -20,6 +20,8 @@ export interface ReceiptEnvelope<E = Record<string, unknown>> {
     sha256: string;
 }
 export declare const RECEIPT_SCHEMA_VERSION: 1;
+/** Deterministic stringify with sorted keys (stable hash basis). */
+export declare function canonicalJson(value: unknown): string;
 export declare function sha256Hex(input: string): string;
 export interface BuildReceiptInput<E> {
     receiptId: string;
@@ -90,6 +92,10 @@ export interface ReviewVerdictEvidence {
     finalizedAt: string;
     /** Bounded summary code/reference for the verdict; never raw assistant text. */
     summaryRef: string | null;
+    /** Where the verdict came from: explicit operator input or extracted from final assistant text. */
+    verdictSource?: "input" | "assistant";
+    /** sha256 of the assistant text the verdict was extracted from, when sourced from the agent. */
+    assistantDigest?: string | null;
 }
 export interface ReviewFailureEvidence {
     /** Machine-actionable reason the review produced no terminal verdict. */
@@ -98,6 +104,46 @@ export interface ReviewFailureEvidence {
     failedAt: string;
     /** Routing hint for the operator/fallback path. */
     fallback: string;
+    /** sha256 of the assistant text examined for a verdict, when one was available. */
+    assistantDigest?: string | null;
+    /** Bounded, whitespace-collapsed assistant summary (never an unbounded transcript dump). */
+    assistantSummary?: string | null;
 }
+/** Pointer back to one superseded child task receipt. */
+export interface PhaseRollupChildPointer {
+    id: string;
+    status: "completed" | "failed" | "aborted" | "merge_failed" | "paused";
+    /** Artifact URI holding the child's full output, when available. */
+    outputUri: string | null;
+    /** Content hash of the child's output artifact, when available. */
+    outputSha256: string | null;
+    /** Hash of the child receipt itself (canonical JSON), for staleness checks. */
+    receiptSha256: string;
+    /**
+     * Per-child ROI accounting carried into the rollup so the aggregate totals
+     * below are recomputable/verifiable from child evidence (not self-reported).
+     * `tokens` is the child's effective token count; cost/cloned are null when
+     * the child reported no such accounting.
+     */
+    tokens: number;
+    costTotal: number | null;
+    clonedTokens: number | null;
+    lowRoi: boolean;
+}
+export interface PhaseRollupEvidence {
+    /** Harness lifecycle boundary this rollup was emitted at. */
+    phase: string;
+    children: PhaseRollupChildPointer[];
+    aggregate: {
+        childCount: number;
+        completed: number;
+        failed: number;
+        totalTokens: number;
+        totalCostTotal: number | null;
+        totalClonedTokens: number | null;
+        lowRoiChildIds: string[];
+    };
+}
+export declare function validatePhaseRollup(e: PhaseRollupEvidence): string[];
 /** Classifications that MUST have a valid `vanish` receipt before the action proceeds. */
 export declare function requiresVanishBeforeAction(classification: RecoveryClassification): boolean;

package/dist/types/harness-control-plane/rpc-adapter.d.ts

@@ -24,6 +24,8 @@ export interface HarnessRpc {
     isLive?(): boolean;
     /** ISO timestamp of the last observed event frame, or null. */
     lastFrameAt?(): string | null;
+    /** Final assistant text from the live session (for review-verdict extraction); null when unavailable. */
+    getLastAssistantText?(): Promise<string | null>;
 }
 export interface AcceptanceResult {
     accepted: boolean;
@@ -54,6 +56,7 @@ export declare class GajaeCodeRpc implements HarnessRpc {
     isLive(): boolean;
     lastFrameAt(): string | null;
     getState(): Promise<RpcStateSnapshot>;
+    getLastAssistantText(): Promise<string | null>;
     sendPrompt(prompt: string): Promise<{
         commandId: string;
         ack: boolean;

package/dist/types/harness-control-plane/state-machine.d.ts

@@ -9,11 +9,16 @@ import type { HarnessLifecycle, NextAllowedAction, PrimitiveResponse, SessionSta
 export declare function isTerminal(lifecycle: HarnessLifecycle): boolean;
 export declare function canTransition(from: HarnessLifecycle, to: HarnessLifecycle): boolean;
 export declare function assertTransition(from: HarnessLifecycle, to: HarnessLifecycle): void;
+export interface NextAllowedActionsOptions {
+    /** Additional live-owner/RPC readiness gate for submit, e.g. rpc-not-idle. */
+    submitUnavailableReason?: string | null;
+}
+export declare function submitUnavailableReason(lifecycle: HarnessLifecycle, ownerLive: boolean, gateReason?: string | null): string | null;
 /**
  * Derive the permitted next actions for a session given its lifecycle and whether
  * a live owner currently holds the lease.
  */
-export declare function nextAllowedActions(lifecycle: HarnessLifecycle, ownerLive: boolean): NextAllowedAction[];
+export declare function nextAllowedActions(lifecycle: HarnessLifecycle, ownerLive: boolean, options?: NextAllowedActionsOptions): NextAllowedAction[];
 export declare function buildStateView(state: SessionState, ownerLive: boolean): SessionStateView;
 /** Build the universal contract response carried by every primitive. */
 export declare function buildResponse<E extends Record<string, unknown>>(state: SessionState, ownerLive: boolean, evidence: E, ok?: boolean): PrimitiveResponse<E>;

package/dist/types/harness-control-plane/types.d.ts

@@ -16,6 +16,14 @@ export type SessionMode = "implement" | "review";
 export type ReviewVerdict = "APPROVE_MERGE_READY" | "REQUEST_CHANGES" | "OWNER_CONFIRMATION_REQUIRED";
 export declare const REVIEW_VERDICTS: readonly ReviewVerdict[];
 export declare function isReviewVerdict(value: unknown): value is ReviewVerdict;
+/**
+ * Extract a single closed-vocabulary review verdict from free-form assistant text.
+ *
+ * Scans for canonical verdict tokens (and accepted aliases) as whole words and returns the
+ * LAST occurrence — the agent's final stated decision wins over any earlier mention. Returns
+ * null when no allowed token is present, so the finalizer fails closed on a missing verdict.
+ */
+export declare function extractReviewVerdict(text: string | null | undefined): ReviewVerdict | null;
 /** Lifecycle states of an operated session. */
 export type HarnessLifecycle = "new" | "started" | "submitted" | "observing" | "recovering" | "validating" | "finalizing" | "completed" | "blocked" | "retired";
 /** Event severities emitted by the owner. */
@@ -27,7 +35,7 @@ export type RiskKind = "normal" | "prompt-not-accepted" | "deleted-worktree" | "
 /** Deterministic recovery classifications. */
 export type RecoveryClassification = "continue" | "send-enter" | "reinject-prompt" | "restart-clean" | "restart-preserve-delta" | "fallback-codex-exec" | "human-check";
 /** Receipt families persisted under the session storage dir. */
-export type ReceiptFamily = "vanish" | "prompt-acceptance" | "validation" | "completion" | "review-verdict" | "review-failure";
+export type ReceiptFamily = "vanish" | "prompt-acceptance" | "validation" | "completion" | "review-verdict" | "review-failure" | "phase-rollup";
 /** The CLI verbs / primitives exposed by `gjc harness <verb>`. */
 export type HarnessVerb = "start" | "submit" | "observe" | "classify" | "recover" | "validate" | "finalize" | "retire" | "events" | "monitor" | "operate";
 /** Submission transports. */
@@ -130,6 +138,10 @@ export interface Observation {
     rpcLive?: boolean;
     /** ISO timestamp of the most recent RPC frame the owner observed, if any. */
     rpcLastFrameAt?: string | null;
+    /** True only when owner/rpc/lifecycle gates indicate a prompt can be submitted now. */
+    readyForSubmit?: boolean;
+    /** Present when readyForSubmit is false; mirrors submit's nextAllowedActions reason. */
+    submitUnavailableReason?: string | null;
 }
 /** Input to the deterministic recovery classifier. */
 export interface ClassifyInput {