@planningo/duul 1.0.0

package/build/schemas/code-review.d.ts

@@ -0,0 +1,514 @@
+import { z } from 'zod';
+export declare const DependenciesSchema: z.ZodObject<{
+    runtime: z.ZodOptional<z.ZodRecord<z.ZodString, z.ZodString>>;
+    dev: z.ZodOptional<z.ZodRecord<z.ZodString, z.ZodString>>;
+}, "strip", z.ZodTypeAny, {
+    runtime?: Record<string, string> | undefined;
+    dev?: Record<string, string> | undefined;
+}, {
+    runtime?: Record<string, string> | undefined;
+    dev?: Record<string, string> | undefined;
+}>;
+export declare const CodeReviewInputSchema: z.ZodObject<{
+    code: z.ZodString;
+    approved_plan: z.ZodString;
+    file_path: z.ZodOptional<z.ZodString>;
+    dependencies: z.ZodOptional<z.ZodObject<{
+        runtime: z.ZodOptional<z.ZodRecord<z.ZodString, z.ZodString>>;
+        dev: z.ZodOptional<z.ZodRecord<z.ZodString, z.ZodString>>;
+    }, "strip", z.ZodTypeAny, {
+        runtime?: Record<string, string> | undefined;
+        dev?: Record<string, string> | undefined;
+    }, {
+        runtime?: Record<string, string> | undefined;
+        dev?: Record<string, string> | undefined;
+    }>>;
+    relevant_code: z.ZodOptional<z.ZodArray<z.ZodObject<{
+        file_path: z.ZodString;
+        code: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        code: string;
+        file_path: string;
+    }, {
+        code: string;
+        file_path: string;
+    }>, "many">>;
+    notes_to_reviewer: z.ZodOptional<z.ZodString>;
+    user_original_request: z.ZodOptional<z.ZodString>;
+    workspace_root: z.ZodOptional<z.ZodString>;
+    project_root: z.ZodOptional<z.ZodString>;
+    working_directories: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+    linked_roots: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+    changed_files: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+    entrypoints: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+    artifact_refs: z.ZodOptional<z.ZodArray<z.ZodObject<{
+        path: z.ZodString;
+        reason: z.ZodString;
+        priority: z.ZodEnum<["high", "medium", "low"]>;
+    }, "strip", z.ZodTypeAny, {
+        path: string;
+        reason: string;
+        priority: "high" | "medium" | "low";
+    }, {
+        path: string;
+        reason: string;
+        priority: "high" | "medium" | "low";
+    }>, "many">>;
+    tracked_only: z.ZodOptional<z.ZodBoolean>;
+    git_head_sha: z.ZodOptional<z.ZodString>;
+    previous_git_head_sha: z.ZodOptional<z.ZodString>;
+    workspace_name: z.ZodOptional<z.ZodString>;
+    setup_script_present: z.ZodOptional<z.ZodBoolean>;
+    run_script_present: z.ZodOptional<z.ZodBoolean>;
+    environment_files_expected: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+    git_diff: z.ZodOptional<z.ZodString>;
+    git_diff_base: z.ZodOptional<z.ZodString>;
+    previous_review_id: z.ZodOptional<z.ZodString>;
+    iteration_count: z.ZodOptional<z.ZodNumber>;
+    max_review_iterations: z.ZodOptional<z.ZodNumber>;
+    reviewer_config: z.ZodOptional<z.ZodObject<{
+        provider: z.ZodOptional<z.ZodEnum<["openai", "anthropic", "google", "openrouter", "compatible"]>>;
+        model: z.ZodOptional<z.ZodString>;
+        base_url: z.ZodOptional<z.ZodString>;
+        api_key: z.ZodOptional<z.ZodString>;
+        temperature: z.ZodOptional<z.ZodNumber>;
+        top_p: z.ZodOptional<z.ZodNumber>;
+    }, "strip", z.ZodTypeAny, {
+        provider?: "openai" | "anthropic" | "google" | "openrouter" | "compatible" | undefined;
+        model?: string | undefined;
+        base_url?: string | undefined;
+        api_key?: string | undefined;
+        temperature?: number | undefined;
+        top_p?: number | undefined;
+    }, {
+        provider?: "openai" | "anthropic" | "google" | "openrouter" | "compatible" | undefined;
+        model?: string | undefined;
+        base_url?: string | undefined;
+        api_key?: string | undefined;
+        temperature?: number | undefined;
+        top_p?: number | undefined;
+    }>>;
+}, "strip", z.ZodTypeAny, {
+    code: string;
+    approved_plan: string;
+    iteration_count?: number | undefined;
+    changed_files?: string[] | undefined;
+    file_path?: string | undefined;
+    relevant_code?: {
+        code: string;
+        file_path: string;
+    }[] | undefined;
+    notes_to_reviewer?: string | undefined;
+    user_original_request?: string | undefined;
+    workspace_root?: string | undefined;
+    project_root?: string | undefined;
+    working_directories?: string[] | undefined;
+    linked_roots?: string[] | undefined;
+    entrypoints?: string[] | undefined;
+    artifact_refs?: {
+        path: string;
+        reason: string;
+        priority: "high" | "medium" | "low";
+    }[] | undefined;
+    tracked_only?: boolean | undefined;
+    git_head_sha?: string | undefined;
+    previous_git_head_sha?: string | undefined;
+    workspace_name?: string | undefined;
+    setup_script_present?: boolean | undefined;
+    run_script_present?: boolean | undefined;
+    environment_files_expected?: string[] | undefined;
+    git_diff?: string | undefined;
+    git_diff_base?: string | undefined;
+    previous_review_id?: string | undefined;
+    max_review_iterations?: number | undefined;
+    reviewer_config?: {
+        provider?: "openai" | "anthropic" | "google" | "openrouter" | "compatible" | undefined;
+        model?: string | undefined;
+        base_url?: string | undefined;
+        api_key?: string | undefined;
+        temperature?: number | undefined;
+        top_p?: number | undefined;
+    } | undefined;
+    dependencies?: {
+        runtime?: Record<string, string> | undefined;
+        dev?: Record<string, string> | undefined;
+    } | undefined;
+}, {
+    code: string;
+    approved_plan: string;
+    iteration_count?: number | undefined;
+    changed_files?: string[] | undefined;
+    file_path?: string | undefined;
+    relevant_code?: {
+        code: string;
+        file_path: string;
+    }[] | undefined;
+    notes_to_reviewer?: string | undefined;
+    user_original_request?: string | undefined;
+    workspace_root?: string | undefined;
+    project_root?: string | undefined;
+    working_directories?: string[] | undefined;
+    linked_roots?: string[] | undefined;
+    entrypoints?: string[] | undefined;
+    artifact_refs?: {
+        path: string;
+        reason: string;
+        priority: "high" | "medium" | "low";
+    }[] | undefined;
+    tracked_only?: boolean | undefined;
+    git_head_sha?: string | undefined;
+    previous_git_head_sha?: string | undefined;
+    workspace_name?: string | undefined;
+    setup_script_present?: boolean | undefined;
+    run_script_present?: boolean | undefined;
+    environment_files_expected?: string[] | undefined;
+    git_diff?: string | undefined;
+    git_diff_base?: string | undefined;
+    previous_review_id?: string | undefined;
+    max_review_iterations?: number | undefined;
+    reviewer_config?: {
+        provider?: "openai" | "anthropic" | "google" | "openrouter" | "compatible" | undefined;
+        model?: string | undefined;
+        base_url?: string | undefined;
+        api_key?: string | undefined;
+        temperature?: number | undefined;
+        top_p?: number | undefined;
+    } | undefined;
+    dependencies?: {
+        runtime?: Record<string, string> | undefined;
+        dev?: Record<string, string> | undefined;
+    } | undefined;
+}>;
+export declare const CodeReviewOutputSchema: z.ZodObject<{
+    verdict: z.ZodEnum<["APPROVE", "REVISE"]>;
+    review_status: z.ZodEnum<["completed", "incomplete"]>;
+    confidence: z.ZodNumber;
+    requires_human_review: z.ZodBoolean;
+    logic_validation: z.ZodString;
+    blocking_issues: z.ZodArray<z.ZodObject<{
+        description: z.ZodString;
+        suggestion: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        description: string;
+        suggestion: string;
+    }, {
+        description: string;
+        suggestion: string;
+    }>, "many">;
+    merge_blockers: z.ZodNullable<z.ZodArray<z.ZodObject<{
+        description: z.ZodString;
+        suggestion: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        description: string;
+        suggestion: string;
+    }, {
+        description: string;
+        suggestion: string;
+    }>, "many">>;
+    non_blocking_suggestions: z.ZodArray<z.ZodString, "many">;
+    vulnerabilities: z.ZodArray<z.ZodObject<{
+        type: z.ZodString;
+        description: z.ZodString;
+        severity: z.ZodEnum<["critical", "high", "medium"]>;
+    }, "strip", z.ZodTypeAny, {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }, {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }>, "many">;
+    optimized_snippet: z.ZodNullable<z.ZodString>;
+    follow_up_todos: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    missing_context: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    evidence_files: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    used_tools: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    tool_exhaustion_reason: z.ZodNullable<z.ZodEnum<["budget", "repeat", "round_limit"]>>;
+    user_original_request_echo: z.ZodNullable<z.ZodString>;
+    symptom_impact: z.ZodNullable<z.ZodObject<{
+        before: z.ZodString;
+        after: z.ZodString;
+        causal_chain: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        before: string;
+        after: string;
+        causal_chain: string;
+    }, {
+        before: string;
+        after: string;
+        causal_chain: string;
+    }>>;
+    symptom_match_notes: z.ZodNullable<z.ZodString>;
+    gates_tripped: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+}, "strip", z.ZodTypeAny, {
+    verdict: "APPROVE" | "REVISE";
+    review_status: "completed" | "incomplete";
+    confidence: number;
+    requires_human_review: boolean;
+    blocking_issues: {
+        description: string;
+        suggestion: string;
+    }[];
+    merge_blockers: {
+        description: string;
+        suggestion: string;
+    }[] | null;
+    non_blocking_suggestions: string[];
+    follow_up_todos: string[] | null;
+    missing_context: string[] | null;
+    evidence_files: string[] | null;
+    used_tools: string[] | null;
+    tool_exhaustion_reason: "budget" | "repeat" | "round_limit" | null;
+    user_original_request_echo: string | null;
+    symptom_impact: {
+        before: string;
+        after: string;
+        causal_chain: string;
+    } | null;
+    symptom_match_notes: string | null;
+    gates_tripped: string[] | null;
+    logic_validation: string;
+    vulnerabilities: {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }[];
+    optimized_snippet: string | null;
+}, {
+    verdict: "APPROVE" | "REVISE";
+    review_status: "completed" | "incomplete";
+    confidence: number;
+    requires_human_review: boolean;
+    blocking_issues: {
+        description: string;
+        suggestion: string;
+    }[];
+    merge_blockers: {
+        description: string;
+        suggestion: string;
+    }[] | null;
+    non_blocking_suggestions: string[];
+    follow_up_todos: string[] | null;
+    missing_context: string[] | null;
+    evidence_files: string[] | null;
+    used_tools: string[] | null;
+    tool_exhaustion_reason: "budget" | "repeat" | "round_limit" | null;
+    user_original_request_echo: string | null;
+    symptom_impact: {
+        before: string;
+        after: string;
+        causal_chain: string;
+    } | null;
+    symptom_match_notes: string | null;
+    gates_tripped: string[] | null;
+    logic_validation: string;
+    vulnerabilities: {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }[];
+    optimized_snippet: string | null;
+}>;
+export declare const CodeReviewMcpOutputSchema: z.ZodObject<{
+    verdict: z.ZodEnum<["APPROVE", "REVISE"]>;
+    review_status: z.ZodEnum<["completed", "incomplete"]>;
+    confidence: z.ZodNumber;
+    requires_human_review: z.ZodBoolean;
+    logic_validation: z.ZodString;
+    blocking_issues: z.ZodArray<z.ZodObject<{
+        description: z.ZodString;
+        suggestion: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        description: string;
+        suggestion: string;
+    }, {
+        description: string;
+        suggestion: string;
+    }>, "many">;
+    merge_blockers: z.ZodNullable<z.ZodArray<z.ZodObject<{
+        description: z.ZodString;
+        suggestion: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        description: string;
+        suggestion: string;
+    }, {
+        description: string;
+        suggestion: string;
+    }>, "many">>;
+    non_blocking_suggestions: z.ZodArray<z.ZodString, "many">;
+    vulnerabilities: z.ZodArray<z.ZodObject<{
+        type: z.ZodString;
+        description: z.ZodString;
+        severity: z.ZodEnum<["critical", "high", "medium"]>;
+    }, "strip", z.ZodTypeAny, {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }, {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }>, "many">;
+    optimized_snippet: z.ZodNullable<z.ZodString>;
+    follow_up_todos: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    missing_context: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    evidence_files: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    used_tools: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+    tool_exhaustion_reason: z.ZodNullable<z.ZodEnum<["budget", "repeat", "round_limit"]>>;
+    user_original_request_echo: z.ZodNullable<z.ZodString>;
+    symptom_impact: z.ZodNullable<z.ZodObject<{
+        before: z.ZodString;
+        after: z.ZodString;
+        causal_chain: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        before: string;
+        after: string;
+        causal_chain: string;
+    }, {
+        before: string;
+        after: string;
+        causal_chain: string;
+    }>>;
+    symptom_match_notes: z.ZodNullable<z.ZodString>;
+    gates_tripped: z.ZodNullable<z.ZodArray<z.ZodString, "many">>;
+} & {
+    review_id: z.ZodString;
+} & {
+    iteration_count: z.ZodNumber;
+    iteration_limit: z.ZodNumber;
+    iteration_limit_reached: z.ZodBoolean;
+} & {
+    token_usage: z.ZodObject<{
+        input_tokens: z.ZodNumber;
+        output_tokens: z.ZodNumber;
+        total_tokens: z.ZodNumber;
+        api_calls: z.ZodNumber;
+        provider: z.ZodString;
+        model: z.ZodString;
+        estimated_cost_usd: z.ZodNullable<z.ZodNumber>;
+        cached_input_tokens: z.ZodOptional<z.ZodNumber>;
+        cache_creation_input_tokens: z.ZodOptional<z.ZodNumber>;
+    }, "strip", z.ZodTypeAny, {
+        provider: string;
+        model: string;
+        input_tokens: number;
+        output_tokens: number;
+        total_tokens: number;
+        api_calls: number;
+        estimated_cost_usd: number | null;
+        cached_input_tokens?: number | undefined;
+        cache_creation_input_tokens?: number | undefined;
+    }, {
+        provider: string;
+        model: string;
+        input_tokens: number;
+        output_tokens: number;
+        total_tokens: number;
+        api_calls: number;
+        estimated_cost_usd: number | null;
+        cached_input_tokens?: number | undefined;
+        cache_creation_input_tokens?: number | undefined;
+    }>;
+}, "strip", z.ZodTypeAny, {
+    iteration_count: number;
+    iteration_limit: number;
+    iteration_limit_reached: boolean;
+    token_usage: {
+        provider: string;
+        model: string;
+        input_tokens: number;
+        output_tokens: number;
+        total_tokens: number;
+        api_calls: number;
+        estimated_cost_usd: number | null;
+        cached_input_tokens?: number | undefined;
+        cache_creation_input_tokens?: number | undefined;
+    };
+    verdict: "APPROVE" | "REVISE";
+    review_status: "completed" | "incomplete";
+    confidence: number;
+    requires_human_review: boolean;
+    blocking_issues: {
+        description: string;
+        suggestion: string;
+    }[];
+    merge_blockers: {
+        description: string;
+        suggestion: string;
+    }[] | null;
+    non_blocking_suggestions: string[];
+    follow_up_todos: string[] | null;
+    missing_context: string[] | null;
+    evidence_files: string[] | null;
+    used_tools: string[] | null;
+    tool_exhaustion_reason: "budget" | "repeat" | "round_limit" | null;
+    user_original_request_echo: string | null;
+    symptom_impact: {
+        before: string;
+        after: string;
+        causal_chain: string;
+    } | null;
+    symptom_match_notes: string | null;
+    gates_tripped: string[] | null;
+    review_id: string;
+    logic_validation: string;
+    vulnerabilities: {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }[];
+    optimized_snippet: string | null;
+}, {
+    iteration_count: number;
+    iteration_limit: number;
+    iteration_limit_reached: boolean;
+    token_usage: {
+        provider: string;
+        model: string;
+        input_tokens: number;
+        output_tokens: number;
+        total_tokens: number;
+        api_calls: number;
+        estimated_cost_usd: number | null;
+        cached_input_tokens?: number | undefined;
+        cache_creation_input_tokens?: number | undefined;
+    };
+    verdict: "APPROVE" | "REVISE";
+    review_status: "completed" | "incomplete";
+    confidence: number;
+    requires_human_review: boolean;
+    blocking_issues: {
+        description: string;
+        suggestion: string;
+    }[];
+    merge_blockers: {
+        description: string;
+        suggestion: string;
+    }[] | null;
+    non_blocking_suggestions: string[];
+    follow_up_todos: string[] | null;
+    missing_context: string[] | null;
+    evidence_files: string[] | null;
+    used_tools: string[] | null;
+    tool_exhaustion_reason: "budget" | "repeat" | "round_limit" | null;
+    user_original_request_echo: string | null;
+    symptom_impact: {
+        before: string;
+        after: string;
+        causal_chain: string;
+    } | null;
+    symptom_match_notes: string | null;
+    gates_tripped: string[] | null;
+    review_id: string;
+    logic_validation: string;
+    vulnerabilities: {
+        type: string;
+        description: string;
+        severity: "high" | "medium" | "critical";
+    }[];
+    optimized_snippet: string | null;
+}>;
+export type CodeReviewInput = z.infer<typeof CodeReviewInputSchema>;
+export type CodeReviewOutput = z.infer<typeof CodeReviewOutputSchema>;
+export type CodeReviewMcpOutput = z.infer<typeof CodeReviewMcpOutputSchema>;

package/build/schemas/code-review.js

@@ -0,0 +1,175 @@
+import { z } from 'zod';
+import { ArtifactRefSchema, ReviewerConfigSchema, IterationMetaOutputSchema, TokenUsageOutputSchema } from './common.js';
+export const DependenciesSchema = z.object({
+    runtime: z
+        .record(z.string(), z.string())
+        .optional()
+        .describe('Runtime package versions, e.g. { "express": "4.18.2" }'),
+    dev: z
+        .record(z.string(), z.string())
+        .optional()
+        .describe('Dev dependency versions, e.g. { "typescript": "5.8.0" }'),
+});
+export const CodeReviewInputSchema = z.object({
+    code: z.string().min(1, 'code must not be empty').describe('The code to review'),
+    approved_plan: z
+        .string()
+        .min(1, 'approved_plan must not be empty')
+        .describe('The previously approved plan this code implements'),
+    file_path: z.string().optional().describe('File path for contextual feedback'),
+    dependencies: DependenciesSchema.optional().describe('Related library version info'),
+    relevant_code: z
+        .array(z.object({
+        file_path: z.string().describe('File path'),
+        code: z.string().describe('Relevant code snippet (types, interfaces, functions, classes)'),
+    }))
+        .optional()
+        .describe('Related code snippets the reviewer should see for full context. ' +
+        'Include type definitions, data models, and surrounding code that the reviewed code depends on.'),
+    notes_to_reviewer: z
+        .string()
+        .optional()
+        .describe('Context or rebuttals from the caller to the reviewer. ' +
+        'Use this to explain codebase-specific facts the reviewer cannot see, ' +
+        'or to respond to blocking issues from a previous round.'),
+    user_original_request: z
+        .string()
+        .max(4000)
+        .optional()
+        .describe("The user's original, unedited problem statement (not paraphrased by the caller). " +
+        'Used by the reviewer to verify the code actually makes the reported symptom go away.'),
+    // --- Workspace-aware scope fields ---
+    workspace_root: z
+        .string()
+        .optional()
+        .describe('Absolute path to the workspace root directory. Preferred over project_root. ' +
+        'When provided, the reviewer gains file exploration tools scoped to this workspace.'),
+    project_root: z
+        .string()
+        .optional()
+        .describe('[DEPRECATED — use workspace_root] Absolute path to the project root directory. ' +
+        'Used as fallback when workspace_root is not provided.'),
+    working_directories: z
+        .array(z.string())
+        .optional()
+        .describe('Subdirectories within workspace_root to restrict file access to. ' +
+        'Acts as a sparse-checkout-like allowlist. If omitted, entire workspace_root is accessible.'),
+    linked_roots: z
+        .array(z.string())
+        .max(5)
+        .optional()
+        .describe('Additional read-only workspace roots the reviewer can access. ' +
+        'Each must be a valid absolute path (3+ depth). Max 5.'),
+    changed_files: z
+        .array(z.string())
+        .optional()
+        .describe('Files changed in this review scope (top-level, separate from project_context)'),
+    entrypoints: z
+        .array(z.string())
+        .optional()
+        .describe('Entry point files the reviewer should start from'),
+    artifact_refs: z
+        .array(ArtifactRefSchema)
+        .max(30)
+        .optional()
+        .describe('References to important files with reason and priority. Max 30.'),
+    tracked_only: z
+        .boolean()
+        .optional()
+        .describe('When true, only git-tracked files are accessible to the reviewer.'),
+    // --- Git metadata ---
+    git_head_sha: z.string().optional().describe('Current git HEAD SHA for this review'),
+    previous_git_head_sha: z.string().optional().describe('Git HEAD SHA from the previous review round. Used with git_head_sha to detect stale context.'),
+    workspace_name: z.string().optional().describe('Name of the workspace (for logging/identification)'),
+    // --- Setup/Run metadata ---
+    setup_script_present: z.boolean().optional().describe('Whether a setup script exists in the workspace'),
+    run_script_present: z.boolean().optional().describe('Whether a run/start script exists in the workspace'),
+    environment_files_expected: z.array(z.string()).optional().describe('Environment files expected but not tracked (e.g. [".env", ".env.local"]). Prevents false positives.'),
+    // --- Git diff ---
+    git_diff: z
+        .string()
+        .optional()
+        .describe('Pre-computed git diff to include in review context. If omitted and workspace_root + changed_files are provided, ' +
+        'the diff is auto-generated.'),
+    git_diff_base: z
+        .string()
+        .optional()
+        .describe('Base ref for auto-generated git diff (e.g. "HEAD", "main"). Default: "HEAD".'),
+    previous_review_id: z
+        .string()
+        .optional()
+        .describe('Response ID from a previous review call. Pass this to maintain reviewer context ' +
+        'across rounds and across phases — the reviewer will remember the plan review ' +
+        'conversation when doing code review.'),
+    // --- Iteration tracking ---
+    iteration_count: z
+        .number()
+        .min(1)
+        .optional()
+        .describe('Current iteration number (1-based). The caller MUST increment this on each call. ' +
+        'When this reaches the iteration limit, the server returns requires_human_review: true.'),
+    max_review_iterations: z
+        .number()
+        .min(1)
+        .max(20)
+        .optional()
+        .describe('Override the maximum iterations for this phase. Default: env MAX_CODE_REVIEW_ITERATIONS or 7.'),
+    // --- Reviewer config ---
+    reviewer_config: ReviewerConfigSchema.optional().describe('Per-request reviewer configuration. Overrides env defaults.'),
+});
+const BlockingIssueSchema = z.object({
+    description: z.string().describe('What the issue is'),
+    suggestion: z.string().describe('How to fix it'),
+});
+const VulnerabilitySchema = z.object({
+    type: z.string().describe('Vulnerability type, e.g. "SQL Injection", "Race Condition"'),
+    description: z.string().describe('Detailed description of the vulnerability'),
+    severity: z.enum(['critical', 'high', 'medium']).describe('Severity level'),
+});
+export const CodeReviewOutputSchema = z.object({
+    verdict: z.enum(['APPROVE', 'REVISE']).describe('Final verdict'),
+    review_status: z.enum(['completed', 'incomplete']).describe('Whether the review was fully completed. "incomplete" means the tool loop was exhausted before the reviewer could finish.'),
+    confidence: z.number().min(0).max(1).describe('Confidence in the verdict (0-1), advisory only'),
+    requires_human_review: z.boolean().describe('Whether a human should review this'),
+    logic_validation: z.string().describe('How accurately the code implements the approved plan'),
+    blocking_issues: z
+        .array(BlockingIssueSchema)
+        .describe('Issues that must be fixed before proceeding'),
+    merge_blockers: z
+        .array(BlockingIssueSchema)
+        .nullable()
+        .describe('Subset of blocking_issues that should block merge. Null if same as blocking_issues.'),
+    non_blocking_suggestions: z
+        .array(z.string())
+        .describe('Optional improvement suggestions'),
+    vulnerabilities: z
+        .array(VulnerabilitySchema)
+        .describe('Security and performance vulnerabilities found'),
+    optimized_snippet: z
+        .string()
+        .nullable()
+        .describe('Codex-suggested optimized code block, or null if not needed'),
+    follow_up_todos: z.array(z.string()).nullable().describe('Follow-up tasks after implementation'),
+    missing_context: z.array(z.string()).nullable().describe('Files or context the reviewer could not access'),
+    evidence_files: z.array(z.string()).nullable().describe('Files the reviewer examined as evidence'),
+    used_tools: z.array(z.string()).nullable().describe('Tool calls made during review'),
+    tool_exhaustion_reason: z.enum(['budget', 'repeat', 'round_limit']).nullable().describe('If review_status is incomplete, the reason why the tool loop was exhausted'),
+    user_original_request_echo: z.string().nullable().describe('Verbatim echo of user_original_request so the reviewer commits to what it was asked to solve. Null only if the caller omitted user_original_request.'),
+    symptom_impact: z
+        .object({
+        before: z.string().describe('Observable symptom the user reported, in their own terms.'),
+        after: z.string().describe('What the user observes now that this code is merged.'),
+        causal_chain: z.string().describe("Why the code change causes 'before' → 'after'."),
+    })
+        .nullable()
+        .describe('How the code changes the user-visible symptom. Null only if user_original_request was not supplied.'),
+    symptom_match_notes: z.string().nullable().describe('If code does NOT clearly address the reported symptom, explain the gap here. Null if fully addressed.'),
+    gates_tripped: z.array(z.string()).nullable().describe('Server-populated list of post-LLM gate names that fired. Reviewer should leave null.'),
+});
+// Extended output with server-added fields (not sent to the reviewer model, used for MCP response)
+export const CodeReviewMcpOutputSchema = CodeReviewOutputSchema
+    .extend({
+    review_id: z.string().describe('Response ID for maintaining reviewer context across rounds. Pass as previous_review_id on the next call.'),
+})
+    .merge(IterationMetaOutputSchema)
+    .merge(TokenUsageOutputSchema);