@lumenflow/core 3.1.3 → 3.2.1

package/dist/wu-schema.js

@@ -0,0 +1,897 @@
+// Copyright (c) 2026 Hellmai Ltd
+// SPDX-License-Identifier: AGPL-3.0-only
+/**
+ * Work Unit YAML Schema
+ *
+ * Zod schema for runtime validation of WU YAML structure.
+ * Provides compile-time type inference and semantic validation.
+ *
+ * Part of WU-1162: Add Zod schema validation to prevent placeholder WU completions
+ * Part of WU-1539: Add BaseWUSchema pattern for create/edit validation
+ *
+ * Schema Architecture (DRY pattern):
+ * - BaseWUSchema: Structural validation only (field types, formats, lengths)
+ * - WUSchema: Extends base + placeholder rejection (for wu:claim, wu:done)
+ * - ReadyWUSchema: Alias for BaseWUSchema (for wu:create, wu:edit)
+ *
+ * @see {@link packages/@lumenflow/cli/src/wu-done.ts} - Consumer (validates spec completeness, uses WUSchema)
+ * @see {@link packages/@lumenflow/cli/src/wu-claim.ts} - Consumer (validates spec completeness, uses WUSchema)
+ * @see {@link packages/@lumenflow/cli/src/wu-create.ts} - Consumer (structural validation, uses ReadyWUSchema)
+ * @see {@link packages/@lumenflow/cli/src/wu-edit.ts} - Consumer (structural validation, uses ReadyWUSchema)
+ * @see {@link packages/@lumenflow/cli/src/validate.ts} - Consumer (CI validation)
+ * @see {@link apps/web/src/lib/llm/schemas/orchestrator.ts} - Pattern reference
+ */
+import { z } from 'zod';
+import { WU_STATUS_GROUPS, WU_DEFAULTS, STRING_LITERALS, WU_EXPOSURE_VALUES, } from './wu-constants.js';
+import { normalizeISODateTime } from './date-utils.js';
+/**
+ * Valid WU status values derived from WU_STATUS constant (DRY principle)
+ * Used for Zod enum validation with improved error messages
+ * Note: Defined as tuple for Zod enum compatibility
+ */
+const VALID_STATUSES = [
+    'todo',
+    'ready',
+    'backlog',
+    'in_progress',
+    'blocked',
+    'done',
+    'completed',
+    'cancelled',
+    'abandoned',
+    'deferred',
+    'closed',
+    'superseded',
+];
+/**
+ * Placeholder sentinel constant
+ *
+ * Used in wu:create template generation and validation.
+ * Single source of truth for placeholder detection (DRY principle).
+ *
+ * @example
+ * // tools/wu-create.ts
+ * description: `${PLACEHOLDER_SENTINEL} Describe the work...`
+ *
+ * @example
+ * // tools/validate.ts
+ * if (doc.description.includes(PLACEHOLDER_SENTINEL)) { error(); }
+ */
+export const PLACEHOLDER_SENTINEL = '[PLACEHOLDER]';
+/**
+ * Minimum description length requirement
+ * Stored as constant for DRY error message generation
+ */
+const MIN_DESCRIPTION_LENGTH = 50;
+/**
+ * WU ID format validation message (DRY principle)
+ * Used across blocks, blocked_by, and ui_pairing_wus fields
+ */
+const WU_ID_FORMAT_MESSAGE = 'Must be WU-XXX format';
+/**
+ * Acceptance criterion error message
+ * Stored as constant for DRY error message generation (sonarjs/no-duplicate-string)
+ */
+const ACCEPTANCE_REQUIRED_MSG = 'At least one acceptance criterion required';
+// =============================================================================
+// WU-1750: NORMALIZATION TRANSFORMS (Watertight YAML validation)
+// =============================================================================
+/**
+ * Regex pattern matching embedded newlines (both literal and escaped)
+ * Handles: "a\nb" (literal newline) and "a\\nb" (escaped backslash-n)
+ */
+const NEWLINE_PATTERN = /\\n|\n/;
+/**
+ * Transform: Normalize string arrays by splitting embedded newlines
+ *
+ * WU-1750: Agents sometimes pass multi-item content as single strings with \n.
+ * This transform auto-repairs: ["a\nb\nc"] → ["a", "b", "c"]
+ *
+ * @example
+ * // Input: ["tools/a.ts\ntools/b.js"]
+ * // Output: ["tools/a.js", "tools/b.js"]
+ */
+const normalizedStringArray = z.array(z.string()).transform((arr) => arr
+    .flatMap((s) => s.split(NEWLINE_PATTERN))
+    .map((s) => s.trim())
+    .filter(Boolean));
+/**
+ * Transform: Normalize description/notes strings by converting escaped newlines
+ *
+ * WU-1750: YAML quoted strings preserve literal \\n as two characters.
+ * This transform converts them to actual newlines: "a\\n\\nb" → "a\n\nb"
+ *
+ * @example
+ * // Input: "Problem:\\n\\n1. First issue"
+ * // Output: "Problem:\n\n1. First issue"
+ */
+const _normalizedMultilineString = z.string().transform((s) => s.replace(/\\n/g, '\n'));
+/**
+ * Refinement: File path cannot contain newlines (post-normalization safety check)
+ *
+ * WU-1750: After normalization, paths should be clean. This catches UnsafeAny edge cases.
+ */
+const filePathItem = z.string().refine((s) => !s.includes('\n') && !s.includes('\\n'), {
+    message: 'File path cannot contain newlines - split into separate array items',
+});
+/**
+ * Normalized code_paths: split embedded newlines + validate each path
+ */
+const normalizedCodePaths = normalizedStringArray.pipe(z.array(filePathItem)).default([]);
+/**
+ * Normalized test paths object: all test arrays normalized
+ */
+const normalizedTestPaths = z
+    .object({
+    manual: normalizedStringArray.optional(),
+    unit: normalizedStringArray.optional(),
+    integration: normalizedStringArray.optional(),
+    e2e: normalizedStringArray.optional(),
+})
+    .optional();
+// =============================================================================
+// BASE FIELD DEFINITIONS (DRY - shared between BaseWUSchema and WUSchema)
+// =============================================================================
+/**
+ * Base description field (structural validation only)
+ * WU-1539: Fixed template string bug (single quotes → function message)
+ * WU-1750: Added normalization of escaped newlines (\\n → actual newlines)
+ */
+const baseDescriptionField = z
+    .string()
+    .min(1, 'Description is required')
+    .transform((s) => s.replace(/\\n/g, '\n')) // WU-1750: Normalize escaped newlines
+    .refine((val) => val.trim().length >= MIN_DESCRIPTION_LENGTH, {
+    // WU-1539 fix: Use function message for dynamic interpolation
+    message: `Description must be at least ${MIN_DESCRIPTION_LENGTH} characters`,
+});
+/**
+ * Strict description field (with placeholder rejection)
+ * Used by wu:claim and wu:done to ensure placeholders are filled
+ * WU-1750: Added normalization of escaped newlines (\\n → actual newlines)
+ */
+const strictDescriptionField = z
+    .string()
+    .min(1, 'Description is required')
+    .transform((s) => s.replace(/\\n/g, '\n')) // WU-1750: Normalize escaped newlines
+    .refine((val) => !val.includes(PLACEHOLDER_SENTINEL), {
+    message: `Description cannot contain ${PLACEHOLDER_SENTINEL} marker`,
+})
+    .refine((val) => val.trim().length >= MIN_DESCRIPTION_LENGTH, {
+    // WU-1539 fix: Use function message for dynamic interpolation
+    message: `Description must be at least ${MIN_DESCRIPTION_LENGTH} characters`,
+});
+/**
+ * Recursive helper: Check all nested values for at least one item
+ * Shared between base and strict acceptance schemas
+ */
+const hasItems = (value) => {
+    if (Array.isArray(value)) {
+        return value.length > 0;
+    }
+    if (typeof value === 'object' && value !== null) {
+        return Object.values(value).some(hasItems);
+    }
+    return false;
+};
+/**
+ * Recursive helper: Check all strings for PLACEHOLDER_SENTINEL
+ * Used only by strict acceptance schema
+ */
+const checkStringsForPlaceholder = (value) => {
+    if (typeof value === 'string') {
+        return !value.includes(PLACEHOLDER_SENTINEL);
+    }
+    if (Array.isArray(value)) {
+        return value.every(checkStringsForPlaceholder);
+    }
+    if (typeof value === 'object' && value !== null) {
+        return Object.values(value).every(checkStringsForPlaceholder);
+    }
+    return true;
+};
+/**
+ * Base acceptance field (structural validation only)
+ * Validates format but allows placeholder markers
+ * WU-1750: Added normalization of embedded newlines in array items
+ */
+const baseAcceptanceField = z.union([
+    // Flat array format (legacy): acceptance: ["item1", "item2"]
+    // WU-1750: Normalize embedded newlines: ["1. a\n2. b"] → ["1. a", "2. b"]
+    normalizedStringArray.pipe(z.array(z.string()).min(1, ACCEPTANCE_REQUIRED_MSG)),
+    // Nested object format (structured): acceptance: { category1: ["item1"], category2: ["item2"] }
+    z.record(z.string(), normalizedStringArray).refine((obj) => Object.values(obj).some(hasItems), {
+        message: ACCEPTANCE_REQUIRED_MSG,
+    }),
+]);
+/**
+ * Strict acceptance field (with placeholder rejection)
+ * Used by wu:claim and wu:done to ensure placeholders are filled
+ * WU-1750: Added normalization of embedded newlines in array items
+ */
+const strictAcceptanceField = z.union([
+    // Flat array format (legacy): acceptance: ["item1", "item2"]
+    // WU-1750: Normalize embedded newlines: ["1. a\n2. b"] → ["1. a", "2. b"]
+    normalizedStringArray
+        .pipe(z.array(z.string()).min(1, ACCEPTANCE_REQUIRED_MSG))
+        .refine((arr) => !arr.some((item) => item.includes(PLACEHOLDER_SENTINEL)), {
+        message: `Acceptance criteria cannot contain ${PLACEHOLDER_SENTINEL} markers`,
+    }),
+    // Nested object format (structured): acceptance: { category1: ["item1"], category2: ["item2"] }
+    z
+        .record(z.string(), normalizedStringArray)
+        .refine((obj) => Object.values(obj).some(hasItems), {
+        message: ACCEPTANCE_REQUIRED_MSG,
+    })
+        .refine((obj) => checkStringsForPlaceholder(obj), {
+        message: `Acceptance criteria cannot contain ${PLACEHOLDER_SENTINEL} markers`,
+    }),
+]);
+/**
+ * Shared field definitions (same for both base and strict schemas)
+ * DRY: Defined once, used in both schema variants
+ */
+const sharedFields = {
+    /** WU identifier (e.g., WU-1162) */
+    id: z.string().regex(/^WU-\d+$/, 'ID must match pattern WU-XXX'),
+    /** Short title describing the work */
+    title: z.string().min(1, 'Title is required'),
+    /** Lane assignment (parent or sub-lane) */
+    lane: z.string().min(1, 'Lane is required'),
+    /** Work type classification */
+    type: z
+        .enum(['feature', 'bug', 'documentation', 'process', 'tooling', 'chore', 'refactor'], {
+        error: 'Invalid type',
+    })
+        .default(WU_DEFAULTS.type),
+    /** Current status in workflow */
+    status: z
+        .enum(VALID_STATUSES, {
+        error: `Invalid status. Valid values: ${VALID_STATUSES.join(', ')}`,
+    })
+        .default(WU_DEFAULTS.status),
+    /** Priority level */
+    priority: z
+        .enum(['P0', 'P1', 'P2', 'P3'], {
+        error: 'Invalid priority',
+    })
+        .default(WU_DEFAULTS.priority),
+    /** Creation date (YYYY-MM-DD) */
+    created: z.string().regex(/^\d{4}-\d{2}-\d{2}$/, 'Created must be YYYY-MM-DD'),
+    /** Files modified by this WU - WU-1750: Normalized to split embedded newlines */
+    code_paths: normalizedCodePaths,
+    /** Test specifications - WU-1750: All test arrays normalized */
+    tests: normalizedTestPaths.default(WU_DEFAULTS.tests),
+    /** Output artifacts (stamps, docs, etc.) - WU-1750: Normalized */
+    artifacts: normalizedStringArray.optional().default(WU_DEFAULTS.artifacts),
+    /** Upstream WU dependencies (informational, legacy field) - WU-1750: Normalized */
+    dependencies: normalizedStringArray.optional().default(WU_DEFAULTS.dependencies),
+    // === Initiative System Fields (WU-1246) ===
+    /** Parent initiative reference (format: INIT-{number} or slug) */
+    initiative: z.string().optional(),
+    /** Phase number within parent initiative */
+    phase: z.number().int().positive().optional(),
+    /** WU IDs that this WU blocks (downstream dependencies) - WU-1750: Normalized + validated */
+    blocks: normalizedStringArray
+        .pipe(z.array(z.string().regex(/^WU-\d+$/, WU_ID_FORMAT_MESSAGE)))
+        .optional(),
+    /** WU IDs that block this WU (upstream dependencies) - WU-1750: Normalized + validated */
+    blocked_by: normalizedStringArray
+        .pipe(z.array(z.string().regex(/^WU-\d+$/, WU_ID_FORMAT_MESSAGE)))
+        .optional(),
+    /** Cross-cutting tags (orthogonal to initiative) - WU-1750: Normalized */
+    labels: normalizedStringArray.optional(),
+    // === End Initiative System Fields ===
+    /**
+     * WU-1683: First-class plan field, symmetric with initiative `related_plan`.
+     * Set via wu:create --plan, wu:edit --plan, or plan:link --id WU-XXX.
+     */
+    plan: z.string().optional(),
+    /**
+     * WU-1833: References to plans, design docs, external specifications
+     * WU-1834: Supports both flat string array AND nested object format for backwards compatibility
+     *
+     * Flat format (WU-1833+):  ['docs/plans/WU-XXX-plan.md']
+     * Nested format (legacy):  [{file: 'docs/path.md', section: 'heading'}]
+     * Mixed format allowed:    ['path.md', {section: 'heading'}]
+     * Bare object (WU-428):    {file: 'docs/path.md', section: 'heading'}
+     */
+    spec_refs: z
+        .union([
+        // Single object format (WU-428 style): {file: '...', section: '...'}
+        z.object({
+            file: z.string().optional(),
+            section: z.string(),
+        }),
+        // Array format (WU-1833+): strings, objects, or mixed
+        z.array(z.union([
+            z.string(), // Flat format: 'docs/path.md'
+            z.object({
+                // Nested format: {file: 'path', section: 'heading'}
+                file: z.string().optional(),
+                section: z.string(),
+            }),
+        ])),
+    ])
+        .optional(),
+    /** Known risks or constraints - WU-1750: Normalized */
+    risks: normalizedStringArray.optional().default(WU_DEFAULTS.risks),
+    /**
+     * Free-form notes - supports string or array (auto-converted to string)
+     * WU-1750: Normalizes escaped newlines (\\n → actual newlines)
+     */
+    notes: z
+        .union([
+        z.string(),
+        z.array(z.string()), // Legacy array format - will be converted
+    ])
+        .optional()
+        .transform((val) => {
+        // Convert array to newline-joined string (legacy format)
+        if (Array.isArray(val)) {
+            return val.filter((s) => s.trim().length > 0).join(STRING_LITERALS.NEWLINE);
+        }
+        // WU-1750: Normalize escaped newlines in string format
+        if (typeof val === 'string') {
+            return val.replace(/\\n/g, '\n');
+        }
+        return val ?? WU_DEFAULTS.notes;
+    }),
+    /** Requires human review before merge */
+    requires_review: z.boolean().optional().default(WU_DEFAULTS.requires_review),
+    /** Locked state (done WUs only) */
+    locked: z.boolean().optional(),
+    /** Completion date (done WUs only) - auto-normalized to ISO datetime */
+    completed_at: z
+        .string()
+        .optional()
+        .transform((val) => normalizeISODateTime(val)),
+    /** Claimed mode (worktree/branch-only/worktree-pr/branch-pr) */
+    claimed_mode: z.enum(['worktree', 'branch-only', 'worktree-pr', 'branch-pr']).optional(),
+    /**
+     * WU-1589: Canonical branch name for this WU claim.
+     *
+     * Set at claim time to record the actual branch used.
+     * Used by defaultBranchFrom() as highest-priority source for branch resolution.
+     * Essential for branch-pr mode where cloud agents may use non-lane-derived branch names.
+     * Cleared on rollback/release/recover when resetting to ready.
+     */
+    claimed_branch: z.string().optional(),
+    /** Assigned agent email */
+    assigned_to: z.string().email().optional(),
+    /** Claim timestamp - auto-normalized to ISO datetime */
+    claimed_at: z
+        .string()
+        .optional()
+        .transform((val) => normalizeISODateTime(val)),
+    /** Block reason (blocked WUs only) */
+    blocked_reason: z.string().optional(),
+    /** Worktree path (claimed WUs only) */
+    worktree_path: z.string().optional(),
+    /** Current active session ID (WU-1438: auto-set on claim, cleared on done) */
+    session_id: z.string().uuid().optional(),
+    /** Agent sessions (issue logging metadata, WU-1231) */
+    agent_sessions: z
+        .array(z.object({
+        session_id: z.string().uuid(),
+        started: z.string().datetime(),
+        completed: z.string().datetime().optional(),
+        agent_type: z.enum([
+            'claude-code',
+            'codex-cli',
+            'cursor',
+            'gemini-cli',
+            'windsurf',
+            'copilot',
+            'other',
+        ]),
+        context_tier: z.union([z.literal(1), z.literal(2), z.literal(3)]),
+        incidents_logged: z.number().int().min(0).default(0),
+        incidents_major: z.number().int().min(0).default(0),
+        artifacts: z.array(z.string()).optional(),
+    }))
+        .optional(),
+    // === Exposure System Fields (WU-1998) ===
+    /**
+     * WU-1998: Exposure level - defines how the WU exposes functionality to users
+     *
+     * Valid values:
+     * - 'ui': User-facing UI changes (pages, components, widgets)
+     * - 'api': API endpoints called by UI or external clients
+     * - 'backend-only': Backend-only changes (no user visibility)
+     * - 'documentation': Documentation changes only
+     *
+     * Optional during transition period, will become required after backlog update.
+     */
+    exposure: z
+        .enum(WU_EXPOSURE_VALUES, {
+        error: `Invalid exposure value. Valid values: ${WU_EXPOSURE_VALUES.join(', ')}`,
+    })
+        .optional(),
+    /**
+     * WU-1998: User journey description for user-facing WUs
+     *
+     * Recommended for exposure: 'ui' and 'api'.
+     * Describes the end-user interaction flow affected by this WU.
+     */
+    user_journey: z.string().optional(),
+    /**
+     * WU-1998: Related UI WUs for backend/API changes
+     *
+     * For WUs with exposure: 'api', this field lists UI WUs that consume the API.
+     * Ensures backend features have corresponding UI coverage.
+     * Each entry must match WU-XXX format.
+     */
+    ui_pairing_wus: normalizedStringArray
+        .pipe(z.array(z.string().regex(/^WU-\d+$/, WU_ID_FORMAT_MESSAGE)))
+        .optional(),
+    /**
+     * WU-2022: Navigation path for UI-exposed features
+     *
+     * For WUs with exposure: 'ui', specifies the route where the feature is accessible.
+     * Used by wu:done to verify that UI features are actually navigable.
+     * Prevents "orphaned code" where features exist but users cannot access them.
+     *
+     * Example: '/dashboard', '/settings/preferences', '/space'
+     */
+    navigation_path: z.string().optional(),
+    // === End Exposure System Fields ===
+    // === Agent-First Approval Fields (WU-2079 → WU-2080) ===
+    /**
+     * WU-2080: Escalation triggers detected for this WU
+     *
+     * Agent-first model: agents auto-approve by default.
+     * Human escalation only when these triggers are detected:
+     * - sensitive_data: Changes to sensitive data handling
+     * - security_p0: P0 security incident or vulnerability
+     * - budget: Budget/resource allocation above threshold
+     * - cross_lane_arch: Cross-lane architectural decision
+     *
+     * Empty array = no escalation needed, agent proceeds autonomously.
+     */
+    escalation_triggers: z
+        .array(z.enum(['sensitive_data', 'security_p0', 'budget', 'cross_lane_arch']))
+        .optional()
+        .default([]),
+    /**
+     * WU-2080: Human escalation required flag
+     *
+     * Auto-set to true when escalation_triggers is non-empty.
+     * When true, wu:done requires human confirmation before completion.
+     */
+    requires_human_escalation: z.boolean().optional().default(false),
+    /**
+     * WU-2080: Email(s) of approvers who signed off
+     *
+     * Auto-populated with claiming agent at wu:claim.
+     * Additional human approvers added when escalation is resolved.
+     */
+    approved_by: z.array(z.string().email()).optional(),
+    /**
+     * WU-2080: Timestamp when approval was granted
+     *
+     * Auto-set at wu:claim for agent auto-approval.
+     * Updated when human escalation is resolved.
+     */
+    approved_at: z
+        .string()
+        .optional()
+        .transform((val) => normalizeISODateTime(val)),
+    /**
+     * WU-2080: Human who resolved escalation (if UnsafeAny)
+     *
+     * Only set when requires_human_escalation was true and resolved.
+     */
+    escalation_resolved_by: z.string().email().optional(),
+    /**
+     * WU-2080: Timestamp when human resolved escalation
+     */
+    escalation_resolved_at: z
+        .string()
+        .optional()
+        .transform((val) => normalizeISODateTime(val)),
+    // Legacy fields (deprecated, kept for backwards compatibility)
+    /** @deprecated Use escalation_triggers instead */
+    requires_cso_approval: z.boolean().optional().default(false),
+    /** @deprecated Use escalation_triggers instead */
+    requires_cto_approval: z.boolean().optional().default(false),
+    /** @deprecated Use escalation_triggers instead */
+    requires_design_approval: z.boolean().optional().default(false),
+    // === End Agent-First Approval Fields ===
+};
+// =============================================================================
+// SCHEMA DEFINITIONS
+// =============================================================================
+/**
+ * Base WU Schema (structural validation only)
+ *
+ * WU-1539: Used by wu:create and wu:edit for fail-fast structural validation.
+ * Allows placeholder markers - only checks field types, formats, and lengths.
+ *
+ * Use case: Validate WU structure at creation/edit time before placeholders are filled.
+ */
+export const BaseWUSchema = z.object({
+    ...sharedFields,
+    description: baseDescriptionField,
+    acceptance: baseAcceptanceField,
+});
+/**
+ * Ready WU Schema (alias for BaseWUSchema)
+ *
+ * WU-1539: Semantic alias for clarity in wu:create and wu:edit.
+ * Same validation as BaseWUSchema - allows placeholders, enforces structure.
+ */
+export const ReadyWUSchema = BaseWUSchema;
+/**
+ * Strict WU Schema (structural + placeholder rejection)
+ *
+ * Validates WU files against LumenFlow requirements:
+ * - No placeholder text in done WUs
+ * - Minimum description length (50 chars)
+ * - Code paths present for non-documentation WUs
+ * - Proper status/lane/type enums
+ *
+ * Used by wu:claim and wu:done to ensure specs are complete.
+ * Provides runtime validation and TypeScript type inference.
+ */
+export const WUSchema = z.object({
+    ...sharedFields,
+    description: strictDescriptionField,
+    acceptance: strictAcceptanceField,
+});
+/**
+ * TypeScript type inferred from schema
+ *
+ * Single source of truth for both runtime validation and compile-time types.
+ * Replaces manual WU interfaces (DRY principle).
+ *
+ * Note: Type inference available in TypeScript via z.infer<typeof WUSchema>
+ * This is a JavaScript file, so the type export is not needed here.
+ *
+ * @typedef {import('zod').z.infer<typeof WUSchema>} WU
+ */
+/**
+ * Validates WU data against strict schema (placeholder rejection)
+ *
+ * Used by wu:claim and wu:done to ensure specs are complete.
+ * Rejects WUs with placeholder markers.
+ *
+ * @param {unknown} data - Parsed YAML data to validate
+ * @returns {z.SafeParseReturnType<WU, WU>} Validation result
+ *
+ * @example
+ * const result = validateWU(yamlData);
+ * if (!result.success) {
+ *   result.error.issues.forEach(issue => {
+ *     console.error(`${issue.path.join('.')}: ${issue.message}`);
+ *   });
+ * }
+ */
+export function validateWU(data) {
+    return WUSchema.safeParse(data);
+}
+/**
+ * Validates WU data against base schema (structural only)
+ *
+ * WU-1539: Used by wu:create and wu:edit for fail-fast structural validation.
+ * Allows placeholder markers - only checks field types, formats, and lengths.
+ *
+ * @param {unknown} data - Parsed YAML data to validate
+ * @returns {z.SafeParseReturnType<WU, WU>} Validation result
+ *
+ * @example
+ * const result = validateReadyWU(yamlData);
+ * if (!result.success) {
+ *   const errors = result.error.issues
+ *     .map(issue => `  • ${issue.path.join('.')}: ${issue.message}`)
+ *     .join('\n');
+ *   die(`WU YAML validation failed:\n\n${errors}`);
+ * }
+ */
+export function validateReadyWU(data) {
+    return ReadyWUSchema.safeParse(data);
+}
+/**
+ * Validates WU spec completeness for done status
+ *
+ * Additional validation beyond schema for WUs marked as done:
+ * - Code paths required for non-documentation WUs
+ * - Locked must be true
+ * - Completed timestamp must be present
+ *
+ * @param {WU} wu - Validated WU data
+ * @returns {{valid: boolean, errors: string[]}} Validation result
+ *
+ * @example
+ * const schemaResult = validateWU(data);
+ * if (schemaResult.success && data.status === 'done') {
+ *   const completenessResult = validateDoneWU(schemaResult.data);
+ *   if (!completenessResult.valid) {
+ *     console.error(completenessResult.errors);
+ *   }
+ * }
+ */
+export function validateDoneWU(wu) {
+    const errors = [];
+    // Check code_paths for non-documentation WUs
+    if (wu.type !== 'documentation' && wu.type !== 'process') {
+        if (!wu.code_paths || wu.code_paths.length === 0) {
+            errors.push('Code paths required for non-documentation WUs');
+        }
+    }
+    // Note: locked and completed_at are set automatically by wu:done
+    // No need to validate them here (they don't exist yet at validation time)
+    return {
+        valid: errors.length === 0,
+        errors,
+    };
+}
+/**
+ * WU-2080: Human escalation email for notifications
+ *
+ * When escalation triggers fire, this email receives notification.
+ * In production, this would integrate with PagerDuty/Slack.
+ */
+const ESCALATION_EMAIL = 'tom@hellm.ai';
+/**
+ * WU-2080: Valid escalation trigger types
+ *
+ * These are the only conditions that require human intervention.
+ * Everything else is auto-approved by agents.
+ */
+export const ESCALATION_TRIGGER_TYPES = [
+    'sensitive_data', // Sensitive data handling changes
+    'security_p0', // P0 security incident
+    'budget', // Budget/resource above threshold
+    'cross_lane_arch', // Cross-lane architectural decision
+];
+/**
+ * WU-2080: Agent-first approval validation
+ *
+ * AGENT-FIRST MODEL: Agents auto-approve by default.
+ * Human escalation only when escalation_triggers is non-empty
+ * AND requires_human_escalation is true AND not yet resolved.
+ *
+ * Returns:
+ * - valid: true if agent can proceed (no unresolved escalation)
+ * - errors: blocking issues requiring human resolution
+ * - warnings: advisory messages (non-blocking)
+ *
+ * @param {object} wu - Validated WU data
+ * @returns {{valid: boolean, errors: string[], warnings: string[]}}
+ */
+export function validateApprovalGates(wu) {
+    const errors = [];
+    const warnings = [];
+    // Agent-first: check for unresolved escalation triggers
+    const triggers = wu.escalation_triggers || [];
+    const requiresEscalation = wu.requires_human_escalation || triggers.length > 0;
+    if (requiresEscalation) {
+        // Check if escalation was resolved by human
+        const resolved = wu.escalation_resolved_by && wu.escalation_resolved_at;
+        if (!resolved) {
+            errors.push(`Human escalation required for: ${triggers.join(', ')}\n` +
+                `   To resolve: Add escalation_resolved_by: "${ESCALATION_EMAIL}" and escalation_resolved_at to WU YAML\n` +
+                `   Or use: pnpm wu:escalate --resolve --id ${wu.id}`);
+        }
+    }
+    // Legacy backwards compatibility: map old fields to new model
+    if (wu.requires_cso_approval || wu.requires_cto_approval || wu.requires_design_approval) {
+        warnings.push('Using deprecated requires_X_approval fields. Migrate to escalation_triggers model.');
+    }
+    return {
+        valid: errors.length === 0,
+        errors,
+        warnings,
+    };
+}
+/**
+ * WU-2080: Detect escalation triggers from WU content
+ *
+ * Analyzes WU metadata to detect conditions requiring human escalation.
+ * Called by wu:claim to auto-set escalation_triggers.
+ *
+ * @param {object} wu - WU data with lane, type, code_paths
+ * @returns {string[]} Array of triggered escalation types
+ */
+export function detectEscalationTriggers(wu) {
+    const triggers = [];
+    const lane = (wu.lane || '').toLowerCase();
+    const codePaths = wu.code_paths || [];
+    // Sensitive data: Changes to user data or auth
+    const sensitivePatterns = ['pii', 'user-data', 'auth', 'credentials'];
+    const touchesSensitive = codePaths.some((p) => sensitivePatterns.some((pat) => p.toLowerCase().includes(pat)));
+    if (touchesSensitive || lane.includes('pii')) {
+        triggers.push('sensitive_data');
+    }
+    // Security P0: Explicit security lane or auth changes
+    if (wu.priority === 'P0' && lane.includes('security')) {
+        triggers.push('security_p0');
+    }
+    return triggers;
+}
+/**
+ * WU-2080: Generate auto-approval metadata for wu:claim
+ *
+ * Called by wu:claim to auto-approve agents within policy.
+ * Sets approved_by and approved_at, detects escalation triggers.
+ *
+ * @param {object} wu - WU data
+ * @param {string} agentEmail - Email of claiming agent
+ * @returns {{approved_by: string[], approved_at: string, escalation_triggers: string[], requires_human_escalation: boolean}}
+ */
+export function generateAutoApproval(wu, agentEmail) {
+    const triggers = detectEscalationTriggers(wu);
+    const now = new Date().toISOString();
+    return {
+        approved_by: [agentEmail],
+        approved_at: now,
+        escalation_triggers: triggers,
+        requires_human_escalation: triggers.length > 0,
+    };
+}
+/**
+ * @deprecated Use detectEscalationTriggers instead
+ * WU-2079: Legacy function for backwards compatibility
+ */
+export function determineRequiredApprovals(wu) {
+    const triggers = detectEscalationTriggers(wu);
+    return {
+        requires_cso_approval: triggers.includes('security_p0') || triggers.includes('sensitive_data'),
+        requires_cto_approval: triggers.includes('cross_lane_arch'),
+        requires_design_approval: false, // Design no longer requires human escalation
+    };
+}
+/**
+ * WU-1811: Validates and normalizes WU YAML data with auto-fixable normalisations
+ *
+ * This function validates the WU YAML schema and applies fixable normalisations:
+ * - Trimming whitespace from string fields
+ * - Normalizing escaped newlines (\\n → \n)
+ * - Splitting embedded newlines in arrays (["a\nb"] → ["a", "b"])
+ *
+ * Returns:
+ * - valid: true if schema validation passes (after normalisations)
+ * - normalized: the normalized data (even if validation fails, partial normalization is returned)
+ * - errors: validation errors if UnsafeAny
+ * - wasNormalized: true if UnsafeAny normalisations were applied
+ *
+ * @param {unknown} data - Parsed YAML data to validate and normalize
+ * @returns {{valid: boolean, normalized: object|null, errors: string[], wasNormalized: boolean}}
+ *
+ * @example
+ * const { valid, normalized, errors, wasNormalized } = validateAndNormalizeWUYAML(yamlData);
+ * if (valid && wasNormalized) {
+ *   // Write normalized data back to YAML file
+ *   writeWU(wuPath, normalized);
+ * }
+ * if (!valid) {
+ *   die(`Validation failed:\n${errors.join('\n')}`);
+ * }
+ */
+export function validateAndNormalizeWUYAML(data) {
+    // First try to parse with schema (which applies normalizations)
+    const result = WUSchema.safeParse(data);
+    if (!result.success) {
+        // Schema validation failed - return errors
+        const errors = result.error.issues.map((issue) => `${issue.path.join('.')}: ${issue.message}`);
+        return {
+            valid: false,
+            normalized: null,
+            errors,
+            wasNormalized: false,
+        };
+    }
+    // Schema passed - check if data was normalized (compare key fields)
+    const normalized = result.data;
+    const original = typeof data === 'object' && data !== null ? data : {};
+    const wasNormalized = detectNormalizationChanges(original, normalized);
+    return {
+        valid: true,
+        normalized,
+        errors: [],
+        wasNormalized,
+    };
+}
+/**
+ * WU-1833: Validate WU spec completeness with advisory warnings
+ *
+ * Provides soft validation that warns (doesn't fail) when recommended fields are missing.
+ * Used by wu:validate command to surface quality issues without blocking workflow.
+ *
+ * Feature and bug WUs should have:
+ * - notes (implementation context, deployment instructions)
+ * - tests.manual (verification steps)
+ * - spec_refs (links to plans, design docs) - for features only
+ *
+ * @param {object} wu - Validated WU data (must pass WUSchema first)
+ * @returns {{warnings: string[]}} Array of warning messages
+ *
+ * @example
+ * const schemaResult = validateWU(data);
+ * if (schemaResult.success) {
+ *   const { warnings } = validateWUCompleteness(schemaResult.data);
+ *   if (warnings.length > 0) {
+ *     console.warn('Quality warnings:');
+ *     warnings.forEach(w => console.warn(`  ⚠️  ${w}`));
+ *   }
+ * }
+ */
+export function validateWUCompleteness(wu) {
+    const warnings = [];
+    // WU-1384: Skip completeness checks for terminal WUs (done, cancelled, etc.)
+    // These are immutable historical records - enforcing completeness is pointless
+    const status = wu.status ?? '';
+    const isTerminal = WU_STATUS_GROUPS.TERMINAL.includes(status);
+    if (isTerminal) {
+        return { warnings };
+    }
+    const type = wu.type || 'feature';
+    // Only check feature and bug WUs - docs/chore/process don't need these
+    const requiresContext = ['feature', 'bug', 'refactor'].includes(type);
+    if (!requiresContext) {
+        return { warnings };
+    }
+    // Check for notes (implementation context)
+    if (!wu.notes || wu.notes.trim().length === 0) {
+        warnings.push(`${wu.id}: Missing 'notes' field. Add implementation context, deployment instructions, or plan links.`);
+    }
+    // Check for manual tests
+    const hasManualTests = wu.tests?.manual && wu.tests.manual.length > 0;
+    if (!hasManualTests) {
+        warnings.push(`${wu.id}: Missing 'tests.manual' field. Add manual verification steps for acceptance criteria.`);
+    }
+    // Check for spec_refs (features should link to plans/specs)
+    // WU-1062: Accepts both repo-relative paths (docs/04-operations/plans/) and
+    // external paths (~/.lumenflow/plans/, $LUMENFLOW_HOME/plans/, lumenflow://plans/)
+    if (type === 'feature') {
+        const specRefs = wu.spec_refs;
+        const hasSpecRefs = !!specRefs && (specRefs.length ?? 0) > 0;
+        if (!hasSpecRefs) {
+            warnings.push(`${wu.id}: Missing 'spec_refs' field. Link to plan file (docs/04-operations/plans/, lumenflow://plans/, or ~/.lumenflow/plans/) for traceability.`);
+        }
+    }
+    return { warnings };
+}
+/**
+ * WU-1811: Detect if normalizations were applied by comparing original and normalized data
+ *
+ * Compares fields that are commonly normalized:
+ * - description (escaped newlines)
+ * - code_paths (embedded newlines split)
+ * - acceptance (embedded newlines split)
+ *
+ * @param {object} original - Original parsed YAML data
+ * @param {object} normalized - Schema-normalized data
+ * @returns {boolean} True if UnsafeAny normalisations were applied
+ */
+function detectNormalizationChanges(original, normalized) {
+    // Compare description (newline normalization)
+    if (original.description !== normalized.description) {
+        return true;
+    }
+    // Compare code_paths (array splitting)
+    const origPaths = Array.isArray(original.code_paths) ? original.code_paths : [];
+    const normPaths = Array.isArray(normalized.code_paths) ? normalized.code_paths : [];
+    if (origPaths.length !== normPaths.length) {
+        return true;
+    }
+    for (let i = 0; i < origPaths.length; i++) {
+        if (origPaths[i] !== normPaths[i]) {
+            return true;
+        }
+    }
+    // Compare acceptance if both are arrays (most common case)
+    if (Array.isArray(original.acceptance) && Array.isArray(normalized.acceptance)) {
+        if (original.acceptance.length !== normalized.acceptance.length) {
+            return true;
+        }
+        for (let i = 0; i < original.acceptance.length; i++) {
+            if (original.acceptance[i] !== normalized.acceptance[i]) {
+                return true;
+            }
+        }
+    }
+    return false;
+}
+//# sourceMappingURL=wu-schema.js.map