perttool 0.1.0 → 0.3.0

package/dist/help/registry.js

@@ -1,33 +1,20 @@
-const agentHelpCommand = {
-    operation: "agent.help",
-    summary: "Bundled offline profileからprovider別のread-only AI agent guidanceを表示します。",
-    syntax: [
-        "Usage: perttool agent help [<provider> [<surface>]]",
-        "  [--level index|quick|detail]",
-        "  [--format text|json]",
-        "  [--color auto|always|never]",
-    ],
-};
-export function getAgentHelpCommandHelp() {
-    return agentHelpCommand;
-}
 const nodes = [
     {
         id: "syntax",
         title: "DSL syntax",
-        summary: "project、resource、milestone、task、gateを記述するgrammar version 1。",
+        summary: "Grammar versions 1, 2, and 3 for declaring project, resource, milestone, task, and gate.",
         quick: [
             {
                 id: "declarations",
                 title: "Declarations",
-                body: "projectを最初に1件置き、続けてresource、milestone、task、gateを記述します。",
+                body: "Place exactly one project first, followed by resource, milestone, task, and gate declarations.",
             },
         ],
         detail: [
             {
                 id: "indentation",
                 title: "Indentation",
-                body: "fieldは2 spaces、estimate/requiresの子fieldは4 spacesです。Tabは使用できません。",
+                body: "Fields use 2 spaces; child fields in estimate and requires use 4 spaces. Tabs are not allowed.",
             },
         ],
         syntax: [
@@ -52,6 +39,7 @@ const nodes = [
             "syntax.gate",
             "syntax.estimate",
             "syntax.velocity",
+            "syntax.temporal",
             "analysis",
             "errors",
         ],
@@ -59,19 +47,19 @@ const nodes = [
     {
         id: "syntax.project",
         title: "Project syntax",
-        summary: "文書先頭にexactly one置くproject declarationです。",
+        summary: "The exactly-one project declaration at the start of a document.",
         quick: [
             {
                 id: "required",
                 title: "Required fields",
-                body: "title、duration_unit、finishが必須です。Point文書では正のvelocityも必須です。",
+                body: "title, duration_unit, and finish are required. Point documents also require a positive velocity.",
             },
         ],
         detail: [
             {
                 id: "version",
                 title: "Grammar version",
-                body: "version省略時は1として扱い、指定する場合もversion 1だけを受理します。",
+                body: "An omitted version is treated as 1. Version 2 adds temporal fields; version 3 additionally accepts exact Fraction Duration.",
             },
         ],
         syntax: [
@@ -83,17 +71,17 @@ const nodes = [
         examples: [
             { id: "minimal", title: "Minimal project", text: "docs/examples/minimal.pert" },
         ],
-        related: ["syntax", "syntax.duration", "syntax.velocity", "syntax.text"],
+        related: ["syntax", "syntax.duration", "syntax.velocity", "syntax.temporal", "syntax.text"],
     },
     {
         id: "syntax.resource",
         title: "Resource syntax",
-        summary: "Renewable resourceと同時利用可能な整数capacityを宣言します。",
+        summary: "Declares a renewable resource and its integer capacity for concurrent use.",
         quick: [
             {
                 id: "capacity",
                 title: "Capacity",
-                body: "titleと1以上のcapacityが必須です。Taskのrequiresからresource IDを参照します。",
+                body: "title and a capacity of at least 1 are required. Tasks reference the resource ID from requires.",
             },
         ],
         detail: [],
@@ -106,12 +94,12 @@ const nodes = [
     {
         id: "syntax.milestone",
         title: "Milestone syntax",
-        summary: "AoA nodeと現在frontierのreached状態を宣言します。",
+        summary: "Declares an AoA node and its reached state in the current frontier.",
         quick: [
             {
                 id: "state",
                 title: "Milestone state",
-                body: "titleが必須です。stateはplannedまたはreachedで、readyは保存せず依存関係から導出します。",
+                body: "title is required. state is planned or reached; ready is derived from dependencies and is not stored.",
             },
         ],
         detail: [],
@@ -119,24 +107,24 @@ const nodes = [
         examples: [
             { id: "minimal", title: "Reached frontier", text: "docs/examples/minimal.pert" },
         ],
-        related: ["syntax", "syntax.task", "syntax.gate", "syntax.tags", "syntax.text"],
+        related: ["syntax", "syntax.task", "syntax.gate", "syntax.temporal", "syntax.tags", "syntax.text"],
     },
     {
         id: "syntax.task",
         title: "Task syntax",
-        summary: "Taskはmilestone間を結ぶAoA edgeです。",
+        summary: "A task is an AoA edge connecting milestones.",
         quick: [
             {
                 id: "required",
                 title: "Required fields",
-                body: "titleと、durationまたはestimateのどちらかexactly oneを指定します。",
+                body: "Specify title and exactly one of duration or estimate.",
             },
         ],
         detail: [
             {
                 id: "resources",
                 title: "Resources",
-                body: "requires blockの全resourceを同時取得し、task完了まで保持します。",
+                body: "All resources in the requires block are acquired together and held until the task completes.",
             },
         ],
         syntax: ["task ID FROM -> TO:", "  title \"...\"", "  duration 1d"],
@@ -151,17 +139,18 @@ const nodes = [
             "syntax.text",
             "syntax.tags",
             "analysis.resources",
+            "syntax.temporal",
         ],
     },
     {
         id: "syntax.gate",
         title: "Gate syntax",
-        summary: "所要時間0のdependency edgeとしてmilestone間を接続します。",
+        summary: "Connects milestones with a zero-duration dependency edge.",
         quick: [
             {
                 id: "reason",
                 title: "Reason",
-                body: "reasonが必須です。Gateはresourceを要求せず、task durationも持ちません。",
+                body: "reason is required. A gate requires no resources and has no task duration.",
             },
         ],
         detail: [],
@@ -178,19 +167,19 @@ const nodes = [
     {
         id: "syntax.estimate",
         title: "PERT estimate syntax",
-        summary: "optimistic、most_likely、pessimisticの三点見積りを記述します。",
+        summary: "Specifies a three-point estimate: optimistic, most_likely, and pessimistic.",
         quick: [
             {
                 id: "order",
                 title: "Ordering constraint",
-                body: "optimistic <= most_likely <= pessimistic、かつpessimistic > 0が必要です。",
+                body: "optimistic <= most_likely <= pessimistic and pessimistic > 0 are required.",
             },
         ],
         detail: [
             {
                 id: "unit",
                 title: "Duration unit",
-                body: "3値は同じsuffixを使い、project duration_unit（day/hour/point）と一致させます。",
+                body: "All three values use the same suffix, matching the project duration_unit (day/hour/point).",
             },
         ],
         syntax: [
@@ -211,34 +200,71 @@ const nodes = [
     {
         id: "syntax.duration",
         title: "Duration syntax",
-        summary: "Finite decimalにproject unit suffix d、h、pのいずれかを付けます。",
+        summary: "An exact Decimal or, in Grammar 3, reduced Fraction followed by d, h, or p.",
         quick: [],
         detail: [],
-        syntax: ["1d", "2.5h", "3p", "0d"],
+        syntax: ["1d", "2.5h", "3p", "1/3d", "0d"],
         examples: [],
         related: ["syntax.estimate", "syntax.velocity", "analysis"],
     },
+    {
+        id: "syntax.temporal",
+        title: "Temporal syntax",
+        summary: "Grammar 2 and 3 add milestone deadline plus task not_before and deadline fields.",
+        quick: [
+            {
+                id: "anchor",
+                title: "Explicit anchor",
+                body: "Any deadline or not_before requires project.as_of. Date and offset date-time values remain distinct, and perttool never invents a clock or host time zone.",
+            },
+        ],
+        detail: [
+            {
+                id: "semantics",
+                title: "Temporal meaning",
+                body: "not_before is a release constraint for a new task start. Deadlines are desired completion or reach times, not dependency edges or hard caps. Mixed valid calendar kinds remain source-valid but may make a projection unavailable.",
+            },
+        ],
+        syntax: [
+            "project ID:",
+            "  version 2|3",
+            "  as_of DATE|OFFSET_DATE_TIME",
+            "milestone ID:",
+            "  deadline DATE|OFFSET_DATE_TIME",
+            "task ID FROM -> TO:",
+            "  not_before DATE|OFFSET_DATE_TIME",
+            "  deadline DATE|OFFSET_DATE_TIME",
+        ],
+        examples: [
+            {
+                id: "temporal-units",
+                title: "Normative temporal examples",
+                text: "docs/examples/temporal-units.md",
+            },
+        ],
+        related: ["syntax.project", "syntax.task", "syntax.milestone", "analysis.temporal", "editing"],
+    },
     {
         id: "syntax.velocity",
         title: "Velocity syntax",
-        summary: "Pointとday/hourを相互換算するproject-wide velocityを記述します。",
+        summary: "Specifies the project-wide velocity for converting between points and days or hours.",
         quick: [
             {
                 id: "forecast",
                 title: "Velocity forecast",
-                body: "duration_unit pointではvelocityが必須です。PERT値はpのまま計算し、換算値をvelocity forecastとして別に返します。",
+                body: "duration_unit point requires velocity. PERT values are calculated in p and converted values are returned separately as a velocity forecast.",
             },
         ],
         detail: [
             {
                 id: "units",
                 title: "Units",
-                body: "20p/10dまたは20p/80hの形式です。dayとhourの関係は暗黙に推測しません。",
+                body: "Use the form 20p/10d or 20p/80h. The relationship between day and hour is not inferred implicitly.",
             },
             {
                 id: "scope",
                 title: "MVP scope",
-                body: "velocityはproject全体で一定です。team別、resource別、期間別velocityはMVP対象外です。",
+                body: "velocity is constant across the project. Per-team, per-resource, and time-varying velocity are outside the MVP scope.",
             },
         ],
         syntax: [
@@ -258,7 +284,7 @@ const nodes = [
     {
         id: "syntax.indentation",
         title: "Indentation",
-        summary: "Structural indentationはASCII spacesだけを使用します。",
+        summary: "Structural indentation uses ASCII spaces only.",
         quick: [],
         detail: [],
         syntax: ["field: 2 spaces", "nested field: 4 spaces"],
@@ -268,13 +294,13 @@ const nodes = [
     {
         id: "syntax.string",
         title: "String syntax",
-        summary: "JSON string literalと同じdouble-quoted textを使用します。",
+        summary: "Uses double-quoted text with the same rules as a JSON string literal.",
         quick: [],
         detail: [
             {
                 id: "unicode",
                 title: "Unicode and escapes",
-                body: "Unicodeは直接記述でき、escapeはJSON形式を使用します。Unpaired surrogateは受理しません。",
+                body: "Unicode may be written directly, and escapes use JSON syntax. Unpaired surrogates are not accepted.",
             },
         ],
         syntax: ["\"text\"", "\"\\u65e5\\u672c\""],
@@ -284,13 +310,13 @@ const nodes = [
     {
         id: "syntax.text",
         title: "Text field syntax",
-        summary: "description、blocked_reason、reasonはStringまたはblock textを取ります。",
+        summary: "description, blocked_reason, and reason accept a String or block text.",
         quick: [],
         detail: [
             {
                 id: "block",
                 title: "Block text",
-                body: "nonblank contentのcommon indentを除去し、paragraph blankと残りのindentを保持します。",
+                body: "Removes the common indent from nonblank content while preserving paragraph blanks and remaining indentation.",
             },
         ],
         syntax: ["  description \"...\"", "  description |", "    first line"],
@@ -300,7 +326,7 @@ const nodes = [
     {
         id: "syntax.tags",
         title: "Tag list syntax",
-        summary: "Bare IdentifierまたはStringを角括弧内へcomma区切りで記述します。",
+        summary: "A comma-separated list of bare Identifiers or Strings in square brackets.",
         quick: [],
         detail: [],
         syntax: ["  tags [alpha, \"two words\"]", "  tags []"],
@@ -310,12 +336,12 @@ const nodes = [
     {
         id: "syntax.comments",
         title: "Comment syntax",
-        summary: "独立行の#以降をCST triviaとして保持します。",
+        summary: "Preserves # and the following text on standalone lines as CST trivia.",
         quick: [
             {
                 id: "standalone",
                 title: "Standalone only",
-                body: "grammar version 1ではinline commentを許可しません。Stringとblock text内の#はcontentです。",
+                body: "Grammar version 1 does not allow inline comments. # within a String or block text is content.",
             },
         ],
         detail: [],
@@ -326,7 +352,7 @@ const nodes = [
     {
         id: "syntax.top-level",
         title: "Top-level declaration syntax",
-        summary: "Projectを文書先頭に置き、各declarationをcolumn 0から開始します。",
+        summary: "Place the project declaration at the start of the document and begin every declaration at column 0.",
         quick: [],
         detail: [],
         syntax: [
@@ -344,24 +370,24 @@ const nodes = [
     {
         id: "analysis",
         title: "Analysis",
-        summary: "PERT/CPM precedence resultとresource scheduleを分離して返します。",
+        summary: "Returns PERT/CPM precedence results separately from the resource schedule.",
         quick: [
             {
                 id: "results",
                 title: "Separated results",
-                body: "precedence makespanはresourceを無視した下限、resource makespanはcapacityを守る決定的heuristic resultです。",
+                body: "The precedence makespan is a lower bound that ignores resources; the resource makespan is a deterministic heuristic result that respects capacity.",
             },
         ],
         detail: [
             {
                 id: "exact",
                 title: "Exact arithmetic",
-                body: "expected、variance、float、makespanはexact Rationalで計算し、displayだけを--precisionで丸めます。Point文書ではvelocity forecastもexact換算します。",
+                body: "expected, variance, float, and makespan use exact Rational arithmetic; only display values are rounded with --precision. Point documents also convert velocity forecasts exactly.",
             },
             {
                 id: "paths",
                 title: "Critical paths",
-                body: "near-critical subgraphとexact driving pathを区別し、path countは列挙上限と独立に返します。",
+                body: "Distinguishes the near-critical subgraph from exact driving paths, and returns the path count independently of the enumeration limit.",
             },
         ],
         syntax: [
@@ -375,29 +401,29 @@ const nodes = [
                 text: "perttool dag analyze docs/examples/parallel.pert",
             },
         ],
-        related: ["analysis.resources", "next"],
+        related: ["analysis.resources", "analysis.temporal", "next"],
     },
     {
         id: "analysis.resources",
         title: "Resource analysis",
-        summary: "Renewable capacityを守る決定的heuristic scheduleです。",
+        summary: "A deterministic heuristic schedule that respects renewable capacity.",
         quick: [
             {
                 id: "capacity",
                 title: "Capacity override",
-                body: "--capacityはwhat-if入力であり、source documentを書き換えません。",
+                body: "--capacity is a what-if input and does not rewrite the source document.",
             },
         ],
         detail: [
             {
                 id: "algorithm",
                 title: "parallel-sgs v1",
-                body: "priority、precedence float、expected duration、task IDの順でcandidateをscanします。optimal=falseです。",
+                body: "Scans candidates by priority, precedence float, expected duration, and task ID. optimal=false.",
             },
             {
                 id: "witness",
                 title: "Resource witness",
-                body: "resource待ちはrelease taskから開始taskへのanalysis-only resource arcで説明し、正本DAGへ保存しません。",
+                body: "Resource waits are explained by analysis-only resource arcs from releasing tasks to starting tasks and are not stored in the authoritative DAG.",
             },
         ],
         syntax: [
@@ -413,47 +439,83 @@ const nodes = [
         ],
         related: ["analysis", "next"],
     },
+    {
+        id: "analysis.temporal",
+        title: "Temporal analysis",
+        summary: "Projects exact calendar schedules and deadline states separately for precedence and heuristic resources.",
+        quick: [
+            {
+                id: "views",
+                title: "Separate views",
+                body: "AnalysisResult v3 retains base analysis and adds temporal precedence, temporal resource, and deadline evaluations. The resource view remains heuristic and optimal=false.",
+            },
+        ],
+        detail: [
+            {
+                id: "availability",
+                title: "Unavailable causes",
+                body: "Missing or incomparable calendar relationships, Point/time conversion limits, completed history, and unresolved blocks remain explicit states and causes rather than guessed values.",
+            },
+            {
+                id: "ranking",
+                title: "Ranking boundary",
+                body: "Deadline facts are informational for Recommendation version 1. NextResult v4 applies not_before only as a separate release gate after recommendation.",
+            },
+        ],
+        syntax: [
+            "perttool dag analyze FILE --format json",
+            "perttool dag next FILE --format json",
+        ],
+        examples: [
+            {
+                id: "temporal-units",
+                title: "Temporal and deadline cases",
+                text: "docs/examples/temporal-units.md",
+            },
+        ],
+        related: ["analysis", "analysis.resources", "next", "syntax.temporal"],
+    },
     {
         id: "next",
         title: "Next tasks",
-        summary: "NextResult.v3で工程上の推奨taskと、active、ready、runnable_now、blocked_now、upcomingを返します。",
+        summary: "Returns NextResult.v4 recommendations, temporal start authority, and active, ready, runnable_now, blocked_now, and upcoming tasks.",
         quick: [
             {
                 id: "classification",
                 title: "Task classification",
-                body: "readyは依存関係上開始可能なplanned task、runnable_nowは現在のactive allocationを差し引いて同時開始できるreadyの部分集合です。",
+                body: "ready consists of planned tasks whose dependencies permit them to start; runnable_now is the subset of ready tasks that can start concurrently after accounting for current active allocations.",
             },
             {
                 id: "recommendation",
                 title: "Recommendation authority",
-                body: "recommendationはready taskをrecommended、allowed、deferred、discouragedへ分類します。recommended setは同時開始可能ですがheuristicであり、optimal=falseです。",
+                body: "recommendation classifies ready tasks as recommended, allowed, deferred, or discouraged. The recommended set can start concurrently, but is heuristic and optimal=false.",
             },
         ],
         detail: [
             {
                 id: "consumer-safety",
                 title: "Machine-readable explanation",
-                body: "--format jsonはcompleteなPerttool.NextResult.v3説明graphを返します。Consumerはschema_versionと各model versionを先に検査し、未知のdecisive semanticsではtaskを自動開始しません。Textはcomplete=falseのsummaryです。",
+                body: "--format json returns a complete Perttool.NextResult.v4 with the unchanged Recommendation version 1 explanation graph and a separate temporal release gate. Consumers validate every identity and do not start when decisive or temporal authority is unknown.",
             },
             {
                 id: "authority-adoption",
                 title: "AI task selection authority",
-                body: "AIは--format jsonのknown Perttool.NextResult.v3、recommendation interface 1、ranking algorithm 1、reason taxonomy 1.0、explanation/expression/description model 1、locale en、completeかつnot-truncatedなtraceだけをnormal start authorityとして使用します。Macroのrecommended work packageを選んでから対応detailを再解析し、recommended subset、またはrecommended set全件にallowedを1件だけ追加した集合を選びます。Unknown version、incomplete trace、PTREC diagnostic、deferred/discouraged selectionでは開始せず停止します。Task stateまたはcapacity変更後は再解析します。",
+                body: "AI uses only a known Perttool.NextResult.v4 from --format json, Recommendation interface 1, ranking algorithm 1, reason taxonomy 1.0, explanation/expression/description model 1, locale en, a complete non-truncated trace, and temporal policy recommendation_v1_plus_release_gate. Start only IDs in startable_recommended_task_ids. Stop for unknown, incomplete, malformed, future, or unavailable authority, PTREC diagnostics, and deferred or discouraged selections. Reanalyze after task-state or capacity changes.",
             },
             {
                 id: "selection",
                 title: "Resource selection",
-                body: "runnable_nowは既存scheduler順で選ぶoperational subset、recommendationはversion付きranking policyで選ぶauthorityであり、同じ集合とは限りません。CLI rendererはrankingを再実装しません。",
+                body: "runnable_now is an operational subset selected by the existing scheduler order, while recommendation is the authority selected by a versioned ranking policy; they are not necessarily the same set. The CLI renderer does not reimplement ranking.",
             },
             {
                 id: "override-validation",
                 title: "Human override validation",
-                body: "Public CoreのvalidateOverrideはcompleteなNextResult.v3と明示requestからPerttool.OverrideDecision.v1を決定的に生成します。これはread-only validationであり、task state、file、Git、networkを変更しません。reason_textとevidenceへsecret、credential、tokenを含めないでください。Override applyとaudit writeのCLIは未実装です。",
+                body: "The public Core validateOverride deterministically produces Perttool.OverrideDecision.v1 from a complete NextResult.v4 and an explicit request, and cannot bypass a future or unavailable temporal release gate. This is read-only validation and does not change task state, files, Git, or the network.",
             },
             {
                 id: "explanation",
                 title: "Upcoming explanation",
-                body: "upcomingは未到達の開始milestoneとunsatisfied incoming edgeを返します。--explain-depthは0..32、default 1です。",
+                body: "upcoming returns unreached starting milestones and unsatisfied incoming edges. --explain-depth is 0..32; the default is 1.",
             },
         ],
         syntax: [
@@ -467,34 +529,36 @@ const nodes = [
                 text: "perttool dag next docs/examples/parallel.pert --format json",
             },
         ],
-        related: ["analysis", "analysis.resources"],
+        related: ["analysis", "analysis.resources", "analysis.temporal"],
     },
     {
         id: "editing",
         title: "Safe editing",
-        summary: "dsl format、task/milestone/resource mutation、atomic batch、dag advanceをpreviewし、検査済み候補を安全にwriteできます。",
+        summary: "Preview document formatting, project/task/gate/milestone/resource mutations, atomic batches, and DAG advancement, then safely write validated candidates.",
         quick: [
             {
                 id: "current-surface",
                 title: "Current surface",
-                body: "project showはproject metadataをread-onlyで返します。dsl formatはplanFormat、project/entity commandとmutation applyはplanMutation、dag advanceはplanAdvanceの再検査済みcandidate、UTF-16 TextEdit、digest、diffをtext/JSONへ投影します。既定はfileを変更しないpreviewです。",
+                body: "project show returns project metadata read-only. project init creates the smallest valid candidate without writing by default. document format projects planFormat; project/entity commands and batch apply project planMutation; and dag advance projects the revalidated candidate from planAdvance, UTF-16 TextEdits, digest, and diff as text/JSON. Direct gate add/set/remove commands and connected atomic batches use the same preview-first mutation path. The default is a preview that does not modify files.",
             },
         ],
         detail: [
             {
                 id: "write-gate",
                 title: "Safe write",
-                body: "--writeは初回read digestとwrite直前digestを照合してatomic replaceし、--expect-digestでcaller lockを追加できます。--outは既存targetを上書きせず新規作成します。--diffはpreview専用です。",
+                body: "--write atomically replaces a file after comparing the initial-read digest with the digest immediately before writing; --expect-digest adds a caller lock. --out creates a new target without overwriting an existing one. --diff is for previews only.",
             },
         ],
         syntax: [
+            "perttool project init PROJECT_ID --title TITLE --duration-unit UNIT --initial-milestone ID --initial-milestone-title TITLE --finish ID [--out PATH]",
             "perttool project show FILE [--format text|json]",
             "perttool project set FILE [--velocity VELOCITY]... [--write [--expect-digest DIGEST] | --out PATH]",
-            "perttool dsl format FILE [--check] [--diff] [--write [--expect-digest DIGEST] | --out PATH]",
+            "perttool document format FILE [--check] [--diff] [--write [--expect-digest DIGEST] | --out PATH]",
             "perttool task add|set|remove|finish ... [--write [--expect-digest DIGEST] | --out PATH]",
+            "perttool gate add|set|remove ... [--write [--expect-digest DIGEST] | --out PATH]",
             "perttool milestone add|set|remove ... [--write [--expect-digest DIGEST] | --out PATH]",
             "perttool resource add|set|remove ... [--write [--expect-digest DIGEST] | --out PATH]",
-            "perttool mutation apply FILE --request REQUEST.json [--write [--expect-digest DIGEST] | --out PATH]",
+            "perttool batch apply FILE --request REQUEST.json [--write [--expect-digest DIGEST] | --out PATH]",
             "perttool dag advance FILE [--diff] [--write [--expect-digest DIGEST] | --out PATH]",
         ],
         examples: [
@@ -514,34 +578,70 @@ const nodes = [
                 text: "perttool task set plan.pert TASK_ID --status active --write --expect-digest sha256:<64 lowercase hex digits>",
             },
         ],
-        related: ["workflows"],
+        related: ["editing.unit-migration", "syntax.temporal", "workflows"],
+    },
+    {
+        id: "editing.unit-migration",
+        title: "Exact unit migration",
+        summary: "Migrates the complete Duration inventory between Point and matching day or hour units.",
+        quick: [
+            {
+                id: "directions",
+                title: "Exact directions",
+                body: "Supported directions are point to day or hour and the matching time unit back to point. A declared or explicit replacement velocity supplies the exact relationship; day-to-hour is never inferred.",
+            },
+        ],
+        detail: [
+            {
+                id: "safety",
+                title: "Preview and safe write",
+                body: "project migrate-unit previews by default, preserves every absolute temporal token, validates one complete candidate, and shares diff, write, out, digest-lock, race, symlink, and post-write checks with ordinary mutations.",
+            },
+            {
+                id: "batch",
+                title: "Batch exclusion and reversibility",
+                body: "Automatic migration is not a batch.apply member. UnitMigrationResult v2 reports exact converted records, grammar and velocity dispositions, qualifications, and whether values and metadata are reversible.",
+            },
+        ],
+        syntax: [
+            "perttool project migrate-unit FILE --to-unit day|hour|point [--replacement-velocity VELOCITY]",
+            "  [--diff] [--write [--expect-digest DIGEST] | --out PATH]",
+        ],
+        examples: [
+            {
+                id: "point-to-day",
+                title: "Preview Point-to-day migration",
+                text: "perttool project migrate-unit plan.pert --to-unit day --diff",
+            },
+        ],
+        related: ["editing", "syntax.duration", "syntax.velocity", "syntax.temporal"],
     },
     {
         id: "mermaid",
         title: "Mermaid conversion",
-        summary: "dag renderとdag importでlossless Mermaid profileを往復でき、plain importの損失を明示できます。",
+        summary: "Round-trip a lossless Mermaid profile with dag render and dag import, and explicitly report plain-import loss.",
         quick: [
             {
                 id: "authority",
                 title: "Profile authority",
-                body: "%% perttool: semantic recordが正規化DSL意味モデルの復元正本で、flowchart node/edgeは人間向けprojectionです。",
+                body: "%% perttool: semantic records are the authoritative source for restoring the normalized DSL semantic model; flowchart nodes and edges are a human-facing projection.",
             },
         ],
         detail: [
             {
                 id: "fidelity",
                 title: "Lossless boundary",
-                body: "losslessはgrammar v1の正規化意味同値です。comment、blank line、field/declaration順、escape spelling、line endingのbyte同一性は対象外です。",
+                body: "Lossless means normalized semantic equivalence for grammar v1. Byte identity for comments, blank lines, field/declaration order, escape spelling, and line endings is out of scope.",
             },
             {
                 id: "integrity",
                 title: "Fail-closed validation",
-                body: "canonical JSON record、record count、metadata/projection SHA-256、node/edge対応を検査します。Profile header検出後の破損をplain importへ黙って降格しません。",
+                body: "Validates canonical JSON records, record count, metadata/projection SHA-256 values, and node/edge correspondence. Corruption detected after a profile header is not silently downgraded to plain import.",
             },
             {
                 id: "availability",
                 title: "Current availability",
-                body: "exportMermaid/importMermaid Coreとdag render/importは利用できます。perttool profileは復元用semantic metadataをfail-closed検査し、plain import/exportはstable PTCNV lossを返します。SVG、JSON targetは後続sliceです。",
+                body: "The exportMermaid/importMermaid Core and dag render/import are available. The perttool profile fail-closed validates restoration metadata, and plain import/export returns stable PTCNV losses. SVG and JSON targets are later slices.",
             },
         ],
         syntax: [
@@ -565,7 +665,7 @@ const nodes = [
     {
         id: "workflows",
         title: "Workflows",
-        summary: "check、analyze、next、preview、Git commitの順で運用します。",
+        summary: "Operate in the order check, analyze, next, preview, and Git commit.",
         quick: [],
         detail: [],
         syntax: [],
@@ -575,28 +675,28 @@ const nodes = [
     {
         id: "errors",
         title: "Diagnostics",
-        summary: "Stable code、source span、局所help topicで問題を報告します。",
+        summary: "Reports problems with stable codes, source spans, and local help topics.",
         quick: [
             {
                 id: "recovery",
                 title: "Error recovery",
-                body: "独立した構文errorは可能な範囲で複数返し、同じinvalid blockの子行は1 error regionとして抑制します。",
+                body: "Returns multiple independent syntax errors where possible, and suppresses child lines in the same invalid block as one error region.",
             },
         ],
         detail: [
             {
                 id: "phases",
                 title: "Phase suppression",
-                body: "parse errorがある文書ではfield/graph validationを実行せず、原因を直す前の派生PTSEM/PTDAG diagnosticを返しません。",
+                body: "For documents with parse errors, field/graph validation does not run, so derived PTSEM/PTDAG diagnostics are not returned before the cause is fixed.",
             },
             {
                 id: "limit",
                 title: "Diagnostic limit",
-                body: "--max-diagnosticsは1..1000、default 100です。上限超過はdiagnostics_truncatedで確認できます。",
+                body: "--max-diagnostics is 1..1000, with a default of 100. diagnostics_truncated indicates when the limit is exceeded.",
             },
         ],
         syntax: [
-            "perttool dsl check FILE --max-diagnostics 20 --format json",
+            "perttool document check FILE --max-diagnostics 20 --format json",
             "perttool dag analyze FILE --max-diagnostics 20 --format json",
             "perttool dag next FILE --max-diagnostics 20 --format json",
         ],
@@ -606,7 +706,7 @@ const nodes = [
     {
         id: "samples",
         title: "Samples",
-        summary: "Repositoryの規範sample IDを案内します。",
+        summary: "Lists the repository's normative sample IDs.",
         quick: [],
         detail: [],
         syntax: [],
@@ -626,7 +726,7 @@ export function getHelp(topicId, level) {
             topicId: null,
             level: "index",
             title: "perttool DSL help",
-            summary: "利用するtopicを選択してください。",
+            summary: "Select a topic to use.",
             sections: [],
             syntax: [],
             examples: [],
@@ -644,7 +744,7 @@ export function getHelp(topicId, level) {
             topicId,
             level,
             title: "Unknown help topic",
-            summary: "指定されたhelp topicは存在しません。",
+            summary: "The specified help topic does not exist.",
             sections: [],
             syntax: [],
             examples: [],