aiterm-mcp 0.29.9 → 0.29.11

package/CHANGELOG.md

@@ -7,6 +7,19 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.29.11] - 2026-09-01
+
+### Added
+
+- Add `agent_steer` to inject an additional message into a running Codex or Grok turn. Idle sessions return `delivery=idle` without sending text so callers can queue a new turn explicitly.
+
+## [0.29.10] - 2026-08-31
+
+### Fixed
+
+- Return `pty_read(agent_transcript:true)` answer text separately as `aiterm.pty-read-result.v1` structured content while preserving the existing human-readable diagnostic suffix.
+- For Grok and Composer, return only the last non-empty assistant message after the last real user row instead of joining tool-use preambles with the final answer.
+
 ## [0.29.9] - 2026-08-31
 
 ### Fixed
@@ -1401,7 +1414,9 @@ prototype (preserved under `prototype/python/` as the porting source and referen
   `ubuntu-latest` for Node 18/20/22, publishing to npm on `v*` tags with
   provenance.
 
-[Unreleased]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.9...HEAD
+[Unreleased]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.11...HEAD
+[0.29.11]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.10...v0.29.11
+[0.29.10]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.9...v0.29.10
 [0.29.9]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.8...v0.29.9
 [0.29.8]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.7...v0.29.8
 [0.29.7]: https://github.com/kitepon/aiterm-mcp/compare/v0.29.6...v0.29.7

package/README.ja.md

@@ -94,7 +94,7 @@ diagnostics、recovery、update、releaseを所有します。このREADMEと[
 
 **言葉でなく実測で:** 記録済み203テストのベンチマークでは、`pty_read` はコンテキストに載るトークンを生ログの **約 7.1 分の 1** に減らす。しかも pass/fail の判定は畳んでも残る。→ [組み込みシェルツールとの使い分け](#組み込みシェルツールとの使い分け)
 
-15ツール: 6つのPTYツール、正規のagent起動入口`agent_launch`、移行用の旧4alias、`agent_configure`、`claude_turn`、`claude_approval`、`diagnostics`。backendはPOSIXのtmux／Windows nativeのpsmuxなので、MCPサーバやAIクライアントが再起動してもsessionは生き残る。
+16ツール: 6つのPTYツール、正規のagent起動入口`agent_launch`、実行中のCodex／Grokを誘導する`agent_steer`、移行用の旧4alias、`agent_configure`、`claude_turn`、`claude_approval`、`diagnostics`。backendはPOSIXのtmux／Windows nativeのpsmuxなので、MCPサーバやAIクライアントが再起動してもsessionは生き残る。
 
 **v0.28.0では実行基盤harnessとmodelを分離した。** harnessはagent loop・認証・hook・session・transcriptを所有し、modelはその上で選ぶ。Cursor Agent CLIでGPT／Claude／Grokを選んでも完了契約はCursor方式のまま。Composerは別harnessではなく、`harness:"grok-cli", model:"grok-composer-2.5-fast"`で表す。旧4起動ツールは同じ実装へ流れる互換alias。
 
@@ -153,7 +153,7 @@ runtime-error store は canonical dotagents config の `collection.enabled: true
 場合だけ収集し、既定OFF、network送信は行いません。tag起点CIのnpm provenance（OIDC Trusted
 Publishing）で公開し、GitHub Release が Official MCP Registry を再登録します。
 
-**状態:** 開発継続中 · 現行公開版 **v0.29.9** · 動作対象は Linux · WSL2 · macOS · Windows ネイティブ · MIT · [変更履歴](CHANGELOG.md)。
+**状態:** 開発継続中 · 現行公開版 **v0.29.11** · 動作対象は Linux · WSL2 · macOS · Windows ネイティブ · MIT · [変更履歴](CHANGELOG.md)。
 
 ### 更新と巻き戻し
 
@@ -287,7 +287,7 @@ Throughline自体が不要である。
 Claude Code を再起動して、接続を確認:
 
 ```bash
-/mcp        # aiterm が connected・15 ツール公開、と出る
+/mcp        # aiterm が connected・16 ツール公開、と出る
 ```
 
 最初のセッション——4 回の呼び出しで、1 個の永続端末:
@@ -328,7 +328,7 @@ MCP クライアントが aiterm を stdio 越しにプログラムから駆動
 
 ```mermaid
 flowchart LR
-    AI["AI / MCP client<br/>(the orchestrator)"] -->|"pty_send · agent_launch · agent_configure · claude_turn · claude_approval<br/>旧launcher alias · diagnostics"| S["aiterm-mcp<br/>stdio MCP · 15 tools"]
+    AI["AI / MCP client<br/>(the orchestrator)"] -->|"pty_send · agent_launch · agent_steer · agent_configure · claude_turn · claude_approval<br/>旧launcher alias · diagnostics"| S["aiterm-mcp<br/>stdio MCP · 16 tools"]
     S -->|"pty_read<br/>token-reduced"| AI
     S -->|"tmux / psmux<br/>send · capture"| P["persistent PTYs<br/>再起動を跨ぐ"]
     P -->|"ssh · docker · repl"| R["nested<br/>remote · container · REPL"]
@@ -409,6 +409,7 @@ aiterm は同じ核心の洞察——端末を出会いの場にする——を
 | `pty_close` | 冪等に閉じ、`closed` / `already_closed`を返す | `session_id` |
 | `pty_list` | セッション一覧（agent行は正規`harness=<id>`と互換`agent=<kind>`を含む） | （なし） |
 | `agent_launch` | harnessとmodelを別軸で選ぶ正規agent起動入口 | `harness`, `prompt?`, `model?`, `reasoning_effort?`, `cwd?`, `write_scope?` |
+| `agent_steer` | 実行中のCodex／Grok turnへtextを差し込む。idleなら送信せず`idle`を返す | `session_id`, `text` |
 | `claude_agent` / `codex_agent` / `grok_agent` / `composer_agent` | deprecated互換alias | 旧launcher引数 |
 | `agent_configure` | 起動中のClaude／Codex／Grok／Composer／Cursorを再起動せずmodel／effort変更 | `session_id`, `model?`, `reasoning_effort?` |
 | `claude_turn` | 相関済みClaude operationをdispatch（issue）または回収（recover） | `action`, `session_id`, `operation_id`, `text?` |
@@ -443,7 +444,7 @@ handoff contextを前置きできる。この任意経路は`throughline >= 0.9.
 `launch_operation_id`とは併用不可で、元sessionのDB所属を変更しない。Throughlineは
 `THROUGHLINE_BIN`、次に`PATH`から解決し、不在・不正・空のexportはPTY作成前に明示失敗する。
 
-エージェントの回答が画面tailより長ければ、`pty_read({ agent_transcript:true })`で再promptなしに全文回収する。Claudeはlaunch相関Stop hook、Codexは通常rollout、Grokは通常session history、Cursorはlaunch IDでbindした通常agent transcriptから同じturnを回収する。
+エージェントの回答が画面tailより長ければ、`pty_read({ agent_transcript:true })`で再promptなしに全文回収する。既存の人間向けcontentは診断suffixを維持し、機械呼出し側は`aiterm.pty-read-result.v1`の`structuredContent.text`から回答本文だけを取得する。Claudeはlaunch相関Stop hook、Codexは通常rollout、Grokは最後の実user行以後の最後の空でないassistantメッセージだけ、Cursorはlaunch IDでbindした通常agent transcriptから同じturnを回収する。
 
 ### 完了検出（5 層）
 

package/README.md

@@ -96,7 +96,7 @@ Aiterm and is not a runtime dependency.
 
 **Measured, not claimed:** in the recorded 203-test benchmark, a `pty_read` puts **~7.1× fewer tokens** in your context than the raw log — and the pass/fail verdict survives the fold. → [When to reach for it vs. the built-in shell](#when-to-reach-for-it-vs-the-built-in-shell)
 
-Fifteen tools: six **PTY tools** — `pty_open` / `pty_send` / `pty_read` / `pty_key` / `pty_close` / `pty_list` — to open, drive, and read one persistent terminal; one canonical **agent launcher**, `agent_launch`, which selects `claude-code`, `codex-cli`, `grok-cli`, or `cursor-cli` as the execution harness; four deprecated launcher aliases kept for migration; `agent_configure`; `claude_turn`; `claude_approval`; and `diagnostics`. The backend is **tmux on POSIX and psmux on native Windows**, so sessions survive even if the MCP server or the AI client restarts.
+Sixteen tools: six **PTY tools** — `pty_open` / `pty_send` / `pty_read` / `pty_key` / `pty_close` / `pty_list` — to open, drive, and read one persistent terminal; one canonical **agent launcher**, `agent_launch`, which selects `claude-code`, `codex-cli`, `grok-cli`, or `cursor-cli` as the execution harness; `agent_steer` for an active Codex or Grok turn; four deprecated launcher aliases kept for migration; `agent_configure`; `claude_turn`; `claude_approval`; and `diagnostics`. The backend is **tmux on POSIX and psmux on native Windows**, so sessions survive even if the MCP server or the AI client restarts.
 
 **v0.28.0 separates the execution harness from the model.** The harness owns the agent loop, authentication, hooks, session, and transcript; `model` is what that harness runs. Cursor Agent CLI can therefore select GPT, Claude, or Grok without changing the completion contract from Cursor hooks to another harness's. Grok Composer is a Grok CLI model preset, not another harness: use `harness: "grok-cli", model: "grok-composer-2.5-fast"`. The old four launcher tools are thin compatibility aliases over the same implementation.
 
@@ -169,7 +169,7 @@ collection is off by default and performs no network I/O. It ships via
 tag-triggered CI with npm provenance (OIDC Trusted Publishing); the GitHub
 Release re-registers the Official MCP Registry entry.
 
-**Status:** actively maintained · current public release **v0.29.9** · runs on Linux · WSL2 · macOS · native Windows (tmux on POSIX, the tmux-CLI-compatible [psmux](https://github.com/psmux/psmux) on native Windows — no WSL required) · MIT · see the [CHANGELOG](CHANGELOG.md).
+**Status:** actively maintained · current public release **v0.29.11** · runs on Linux · WSL2 · macOS · native Windows (tmux on POSIX, the tmux-CLI-compatible [psmux](https://github.com/psmux/psmux) on native Windows — no WSL required) · MIT · see the [CHANGELOG](CHANGELOG.md).
 
 ### Update and rollback
 
@@ -316,7 +316,7 @@ The only edits to the captures above are the two `⋮` lines (a long head/tail r
 Restart Claude Code, then verify the connection:
 
 ```bash
-/mcp        # aiterm should show as connected, exposing 15 tools
+/mcp        # aiterm should show as connected, exposing 16 tools
 ```
 
 Your first session — four calls, one persistent terminal:
@@ -357,7 +357,7 @@ The terminal is real and shared, so a human *can* jump in ([A human can watch](#
 
 ```mermaid
 flowchart LR
-    AI["AI / MCP client<br/>(the orchestrator)"] -->|"pty_send · agent_launch · agent_configure · claude_turn · claude_approval<br/>legacy launcher aliases · diagnostics"| S["aiterm-mcp<br/>stdio MCP · 15 tools"]
+    AI["AI / MCP client<br/>(the orchestrator)"] -->|"pty_send · agent_launch · agent_steer · agent_configure · claude_turn · claude_approval<br/>legacy launcher aliases · diagnostics"| S["aiterm-mcp<br/>stdio MCP · 16 tools"]
     S -->|"pty_read<br/>token-reduced"| AI
     S -->|"tmux / psmux<br/>send · capture"| P["persistent PTYs<br/>survive restarts"]
     P -->|"ssh · docker · repl"| R["nested<br/>remote · container · REPL"]
@@ -440,6 +440,7 @@ On top of that sits a productized layer a raw tmux bridge doesn't have: **token-
 | `pty_close` | Close idempotently; return `closed` / `already_closed` | `session_id` |
 | `pty_list` | List sessions (agent rows carry canonical `harness=<id>` plus compatibility `agent=<kind>`) | (none) |
 | `agent_launch` | Canonical agent launch; harness and model are independent | `harness`, `prompt?`, `model?`, `reasoning_effort?`, `cwd?`, `write_scope?` |
+| `agent_steer` | Inject text into the active Codex or Grok turn; return `idle` without sending when no turn is active | `session_id`, `text` |
 | `claude_agent` / `codex_agent` / `grok_agent` / `composer_agent` | Deprecated compatibility aliases | legacy launcher arguments |
 | `agent_configure` | Change model/effort in a running Claude, Codex, Grok, Composer, or Cursor session without restarting it | `session_id`, `model?`, `reasoning_effort?` |
 | `claude_turn` | Issue (dispatch-only) or recover one correlated Claude operation | `action`, `session_id`, `operation_id`, `text?` |
@@ -475,7 +476,7 @@ cannot be combined with `launch_operation_id`, and leaves the source session's d
 unchanged. Throughline is resolved through `THROUGHLINE_BIN` and then `PATH`; a missing or invalid
 export fails before the PTY exists instead of silently launching clean.
 
-When an agent's answer is longer than the on-screen tail (pane height ≈ 24 lines), callers recover it in full with `pty_read({ agent_transcript: true })`. It returns the most recently completed turn's final assistant message in plain text with no re-prompting. Claude reads the bounded owner-only result captured by the launch-correlated Stop hook and verifies its digest/byte count; it never reads Claude's private transcript. Durable machine callers should use `claude_turn`: `issue` sends once, `recover` never sends, `pending` is distinct from unsafe or malformed state, and only `completed` carries the exact verified `raw_output`. Codex uses the normal rollout transcript's `task_complete.turn_id`; Grok/Composer use their normal session history after the last real user row; Cursor uses the normal agent transcript bound to the launch ID and current turn. Missing or ambiguous attribution remains an explicit error.
+When an agent's answer is longer than the on-screen tail (pane height ≈ 24 lines), callers recover it in full with `pty_read({ agent_transcript: true })`. It returns the most recently completed turn's final assistant message in plain text with no re-prompting. The existing human-readable content keeps its diagnostic suffix; machine callers read the answer alone from `structuredContent.text` in `aiterm.pty-read-result.v1`. Claude reads the bounded owner-only result captured by the launch-correlated Stop hook and verifies its digest/byte count; it never reads Claude's private transcript. Durable machine callers should use `claude_turn`: `issue` sends once, `recover` never sends, `pending` is distinct from unsafe or malformed state, and only `completed` carries the exact verified `raw_output`. Codex uses the normal rollout transcript's `task_complete.turn_id`; Grok/Composer return the last non-empty assistant message after the last real user row, excluding tool-use preambles; Cursor uses the normal agent transcript bound to the launch ID and current turn. Missing or ambiguous attribution remains an explicit error.
 
 ### Completion detection (5 layers)
 

package/dist/core.js

@@ -1994,8 +1994,8 @@ async function settlePublishedClaudeCompletionMarker(meta, marker) {
     }
     return active;
 }
-/** agent harness の構造化 transcript から直近完了ターンの最終回答を読む。 */
-export async function readAgentTranscript(name, o = {}) {
+/** agent harness の構造化 transcript から直近完了ターンの最終回答と相関情報を読む。 */
+export async function readAgentTranscriptResult(name, o = {}) {
     const meta = loadAgentMetadata(name);
     const operationId = o.operation_id == null ? null : validateOperationId(o.operation_id);
     if (operationId && meta.kind !== "claude") {
@@ -2050,7 +2050,18 @@ export async function readAgentTranscript(name, o = {}) {
         done?.operation_id ? `operation_id=${done.operation_id}` : null,
         `raw_chars=${rawChars}`,
     ].filter(Boolean).join(" ");
-    return `${body}\n${outputMeta} [${transcriptMeta}]`;
+    return {
+        text: body,
+        display: `${body}\n${outputMeta} [${transcriptMeta}]`,
+        vendor: meta.kind,
+        turn_id: turnId,
+        harness: agentHarness(meta.kind),
+        raw_chars: rawChars,
+    };
+}
+/** 人間向け互換表示を維持する。機械利用は readAgentTranscriptResult の text を使う。 */
+export async function readAgentTranscript(name, o = {}) {
+    return (await readAgentTranscriptResult(name, o)).display;
 }
 function inferAgentFrontend(name, meta, screen) {
     const fg = paneCurrentCommand(name);
@@ -2837,6 +2848,33 @@ export async function dispatchAgentTurn(name, text, o = {}) {
         submit_residue: residue.residue,
     };
 }
+export async function steerAgentTurn(name, text) {
+    assertSessionName(name);
+    const meta = loadAgentMetadata(name);
+    if (!["codex", "grok", "composer"].includes(meta.kind)) {
+        throw new AitermError("agent_steer はCodex/Grok agent sessionだけで使用できます", 2);
+    }
+    const receipt = {
+        schema: "aiterm.agent-steer.v1",
+        session_id: meta.aiterm_session,
+        launch_id: meta.launch_id,
+        vendor: meta.kind,
+        harness: agentHarness(meta.kind),
+    };
+    if (!isAgentTuiBusy(meta.kind, captureScreen(name, AGENT_TUI_READY_LINES)))
+        return { ...receipt, delivery: "idle" };
+    send(name, text, {
+        enter: false,
+        force: true,
+        raw: false,
+        mark: false,
+        rtk: false,
+        bracketedPaste: true,
+    });
+    await sleep(AGENT_SUBMIT_DELAY_MS);
+    sendKey(name, "Enter");
+    return { ...receipt, delivery: "steered" };
+}
 export async function runClaudeOperation({ session_id: name, action, operation_id: operationIdInput, text, }) {
     assertSessionName(name);
     if (action !== "issue" && action !== "recover") {

package/dist/harnesses/grok.js

@@ -283,7 +283,7 @@ export function grokFooterHasConfiguration(screen, model, effort) {
         return modelLabel !== null || effort !== null;
     });
 }
-// 最後のuser発話以降のassistantメッセージをchat_history.jsonlから抽出する。
+// 最後のuser発話以降に確定した最後のassistantメッセージをchat_history.jsonlから抽出する。
 export function grokTranscriptText(meta, readTranscriptLines, transcriptUnavailable) {
     if (!meta.grok_home || !meta.vendor_session_id)
         transcriptUnavailable();
@@ -307,11 +307,12 @@ export function grokTranscriptText(meta, readTranscriptLines, transcriptUnavaila
             // 外部 transcript の壊れた1行は残りの完結行を読む妨げにしない。
         }
     }
-    return records
+    const replies = records
         .slice(lastUser + 1)
         .filter((record) => record?.type === "assistant" && typeof record?.content === "string")
-        .map((record) => record.content)
-        .join("\n");
+        .map((record) => record.content.trim())
+        .filter(Boolean);
+    return replies.at(-1) ?? "";
 }
 export function createGrokAgentMetadata(kind, name, cwd, initialPrompt, authPath, writeScope, lineageContext) {
     const launchId = randomBytes(16).toString("hex");

package/dist/index.js

@@ -203,6 +203,33 @@ server.registerTool("pty_send", {
         return fail(e);
     }
 });
+server.registerTool("agent_steer", {
+    description: "実行中のCodex/Grok agentへ追加メッセージを差し込み、現在のターンを誘導する。" +
+        "独立した次ターンを始める用途ではなく、idle時は文字を送らずdelivery=idleを返す。",
+    inputSchema: {
+        session_id: z.string(),
+        text: z.string().describe("現在のターンへ追加する文字列。UTF-8で最大64KiB"),
+    },
+    outputSchema: {
+        schema: z.literal("aiterm.agent-steer.v1"),
+        session_id: z.string(),
+        launch_id: z.string(),
+        vendor: z.enum(["codex", "grok", "composer"]),
+        harness: z.enum(["codex-cli", "grok-cli"]),
+        delivery: z.enum(["steered", "idle"]),
+    },
+}, async ({ session_id, text }) => {
+    try {
+        const receipt = await core.steerAgentTurn(session_id, text);
+        return {
+            content: [{ type: "text", text: `${receipt.delivery} ${receipt.session_id}` }],
+            structuredContent: receipt,
+        };
+    }
+    catch (e) {
+        return fail(e);
+    }
+});
 server.registerTool("pty_read", {
     description: "セッションの出力をトークン削減して読む（既定は前回読取位置からの増分）。" +
         "削減: 制御文字除去 / 反復圧縮 / head+tail 折りたたみ＋復元ヒント＋メタ併記。" +
@@ -239,16 +266,39 @@ server.registerTool("pty_read", {
             .nullish()
             .describe("Claude operationの期待ID。agent_transcript:true時だけ指定し、古い別operationの結果を拒否する"),
     },
+    outputSchema: {
+        schema: z.literal("aiterm.pty-read-result.v1"),
+        mode: z.enum(["terminal", "agent_transcript"]),
+        session_id: z.string(),
+        text: z.string(),
+        vendor: z.enum(["claude", "codex", "grok", "composer", "cursor"]).nullable(),
+        turn_id: z.string().nullable(),
+        harness: z.enum(["claude-code", "codex-cli", "grok-cli", "cursor-cli"]).nullable(),
+        raw_chars: z.number().int().nonnegative().nullable(),
+    },
 }, async ({ session_id, wait, until, until_regex, timeout, screen, full, lines, line_range, raw, rtk, agent_transcript, operation_id }) => {
     try {
         if (agent_transcript) {
             if (screen || full || rtk || wait || line_range != null) {
                 throw new Error("agent_transcript:true は screen / full / rtk / line_range / wait と併用できません。lines のみ指定できます。");
             }
-            return ok(await core.readAgentTranscript(session_id, {
+            const result = await core.readAgentTranscriptResult(session_id, {
                 lines: lines ?? null,
                 operation_id: operation_id ?? null,
-            }));
+            });
+            return {
+                content: [{ type: "text", text: result.display }],
+                structuredContent: {
+                    schema: "aiterm.pty-read-result.v1",
+                    mode: "agent_transcript",
+                    session_id,
+                    text: result.text,
+                    vendor: result.vendor,
+                    turn_id: result.turn_id,
+                    harness: result.harness,
+                    raw_chars: result.raw_chars,
+                },
+            };
         }
         if (operation_id != null)
             throw new Error("operation_idはagent_transcript:true時だけ指定できます");
@@ -279,7 +329,19 @@ server.registerTool("pty_read", {
             raw,
             rtk,
         });
-        return ok(out);
+        return {
+            content: [{ type: "text", text: out }],
+            structuredContent: {
+                schema: "aiterm.pty-read-result.v1",
+                mode: "terminal",
+                session_id,
+                text: out,
+                vendor: null,
+                turn_id: null,
+                harness: null,
+                raw_chars: null,
+            },
+        };
     }
     catch (e) {
         return fail(e);

package/docs/DESIGN.md

@@ -25,6 +25,7 @@ project／user環境を置換せず、launch相関と完了回収に必要なsta
 agent turnは常に非ブロックdispatchである。receiptの`event_cursor`がturn境界、`wait_process`が
 platform nativeな別process起動情報を返す。waiterは純readerで、親のforeground turnを塞がない。
 回答はharness所有transcriptから同じturnへ相関して回収し、欠落・曖昧・timeout時にpromptを再送しない。
+`agent_steer`は実行中のCodex／Grok turnへ追加textを差し込み、idleなら送信せず状態を返す。
 
 ## Layer ownership
 

package/docs/RELEASE.md

@@ -26,7 +26,7 @@ npm pack --dry-run
 npm run mcpb:build
 ```
 
-MCPBのstaged serverでversion、15 tools、stderr 0、必要なruntime JavaScriptの同梱を確認する。
+MCPBのstaged serverでversion、16 tools、stderr 0、必要なruntime JavaScriptの同梱を確認する。
 
 ## Mainと公開
 
@@ -44,7 +44,7 @@ CI callerは同じrepositoryの`./.github/workflows/product-full-ci.yml`だけ
 公式npm packageを隔離またはglobal installし、次を確認する。
 
 - `aiterm-mcp`、`aiterm-wait`、`aiterm-runtime-errors`の3 bins。
-- MCP initializeのversion、15 tools、stderr 0。
+- MCP initializeのversion、16 tools、stderr 0。
 - POSIXはtmux、Windows nativeはpsmux 3.3.8以上とPowerShell 7。
 - 変更に触れたharnessの起動、non-blocking dispatch、wait outcome、transcript回収、`pty_close`後の残骸ゼロ。
 - Official Registryが`io.github.kitepon/aiterm-mcp`の同じversionをactive／latestとして返す。

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "aiterm-mcp",
-  "version": "0.29.9",
+  "version": "0.29.11",
   "mcpName": "io.github.kitepon/aiterm-mcp",
   "description": "Persistent terminal MCP with one harness-based launcher for Claude Code, Codex CLI, Grok CLI, and Cursor Agent CLI, plus durable PTYs for SSH, containers, and REPLs.",
   "keywords": [