blun-king-cli 9.1.505 → 9.1.507

package/CHANGELOG.md

@@ -1,5 +1,17 @@
 # Changelog
 
+## 9.1.507 - 2026-08-30
+
+- Instructs the Read tool to issue independent file and range reads together in one model response when several inputs may be useful.
+- Reduces avoidable model round trips over large work contexts without changing Read execution, paging, permissions, or result handling.
+- Adds a focused regression and keeps all TodoList, Telegram queue, bot-priority, tool-restoration, and package tests green.
+
+## 9.1.506 - 2026-08-30
+
+- Removes completed TodoList entries from the visible TUI immediately, matching the already-pruned persistent Todo store while retaining every unfinished item.
+- Runs automatic TodoList maintenance with a dedicated 546-character system prompt instead of appending the full work prompt measured at roughly 19,000 characters.
+- Preserves the full work prompt and ordinary tool catalogue on the following step, with regressions covering Todo lifecycle, maintenance isolation, and tool restoration.
+
 ## 9.1.505 - 2026-08-30
 
 - Preserves Telegram messages that arrive after a TUI claims channel ownership but before the delayed channel activation finishes.

package/LIESMICH.txt

@@ -9,7 +9,7 @@ Installation
 ------------
 Die geprüfte Version exakt global installieren:
 
-  npm install -g blun-king-cli@9.1.505
+  npm install -g blun-king-cli@9.1.507
 
 Start
 -----
@@ -65,6 +65,17 @@ geschriebenen Entwurf zu löschen. Das gilt auch bei
 Autovervollständigung, Geistervorschlägen, Bash-Eingabe und einer offenen
 Mehrzeileneingabe.
 
+Version 9.1.507 fordert das Modell im Read-Werkzeug ausdruecklich dazu auf,
+voneinander unabhaengige Datei- und Bereichslesungen in einer Modellantwort zu
+buendeln. Dadurch entfallen vermeidbare Modellrunden, ohne Ausfuehrung,
+Seitennavigation, Berechtigungen oder Ergebnisbehandlung des Werkzeugs zu aendern.
+
+Version 9.1.506 entfernt erledigte TodoList-Punkte sofort aus der sichtbaren
+TUI und verwendet fuer den automatischen Wartungszug einen eigenstaendigen,
+546 Zeichen langen Systemprompt statt des rund 19.000 Zeichen langen
+Arbeits-Systemprompts. Der nachfolgende Arbeitsschritt behaelt weiterhin den
+vollstaendigen Prompt und Werkzeugkatalog.
+
 Version 9.1.504 startet eine automatisch eingereihte Telegram-Nachricht im
 Leerlauf als neuen Modellzug. Sie wartet damit nicht mehr auf einen bereits
 laufenden Zug, der im Leerlauf definitionsgemaess nie entstehen kann.

package/README.md

@@ -9,7 +9,7 @@ Voraussetzung ist Node.js 24.15 oder neuer. Die geprüfte Version wird exakt
 installiert:
 
 ```powershell
-npm install -g blun-king-cli@9.1.505
+npm install -g blun-king-cli@9.1.507
 ```
 
 ## Reproduzierbares Staging und Packen
@@ -81,6 +81,17 @@ konkurrierenden Start zurückgewiesen und bleiben an der Spitze der
 FIFO-Warteschlange. Für diesen normalen Wartestatus erscheint kein
 `turn.agent_busy`-Fehler.
 
+Version 9.1.507 fordert das Modell im Read-Werkzeug ausdrücklich dazu auf,
+voneinander unabhängige Datei- und Bereichslesungen in einer Modellantwort zu
+bündeln. Dadurch entfallen vermeidbare Modellrunden, ohne Ausführung,
+Seitennavigation, Berechtigungen oder Ergebnisbehandlung des Werkzeugs zu ändern.
+
+Version 9.1.506 entfernt erledigte TodoList-Punkte sofort aus der sichtbaren
+TUI und verwendet fuer den automatischen Wartungszug einen eigenstaendigen,
+546 Zeichen langen Systemprompt statt des rund 19.000 Zeichen langen
+Arbeits-Systemprompts. Der nachfolgende Arbeitsschritt behaelt weiterhin den
+vollstaendigen Prompt und Werkzeugkatalog.
+
 Version 9.1.504 startet eine automatisch eingereihte Telegram-Nachricht im
 Leerlauf als neuen Modellzug. Sie wartet damit nicht mehr auf einen bereits
 laufenden Zug, der im Leerlauf definitionsgemäß nie entstehen kann.

package/blun.mjs

@@ -259434,7 +259434,7 @@ var init_read = __esmMin((() => {
 		MAX_BYTES_KB: MAX_BYTES / 1024,
 		MAX_LINE_LENGTH
 	});
-	read_default = "Read a known UTF-8 text file. Invalid paths error; directories are unsupported. Relative paths use the working directory; outside paths must be absolute. Use Glob for unknown names, Grep for content, and ReadMediaFile for media.\n\nEach call returns at most {{ MAX_LINES }} lines or {{ MAX_BYTES_KB }} KB and truncates lines beyond {{ MAX_LINE_LENGTH }} characters. Page large files with 1-based `line_offset` and `n_lines`; omit `n_lines` for the cap. Negative `line_offset` reads from the end, up to {{ MAX_LINES }} lines. A capped result names the exact next `line_offset`.\n\nSensitive files such as `.env` and credential stores are refused; templates and public keys remain readable. Only UTF-8 text without NUL bytes is accepted; use Bash or an MCP tool for other binaries.\n\nOutput is `<line-number>\\t<content>` per line. The trailing `<system>...</system>` block is status, not file content. Pure CRLF appears as LF and is preserved by Edit. Mixed or lone carriage returns appear as `\\r` and require exact `Edit.old_string` escapes.";
+	read_default = "Read a known UTF-8 text file. Invalid paths error; directories are unsupported. Relative paths use the working directory; outside paths must be absolute. Use Glob for unknown names, Grep for content, and ReadMediaFile for media.\n\nEach call returns at most {{ MAX_LINES }} lines or {{ MAX_BYTES_KB }} KB and truncates lines beyond {{ MAX_LINE_LENGTH }} characters. Page large files with 1-based `line_offset` and `n_lines`; omit `n_lines` for the cap. Negative `line_offset` reads from the end, up to {{ MAX_LINES }} lines. A capped result names the exact next `line_offset`. Issue independent Read calls together in one response when several files or ranges may be useful.\n\nSensitive files such as `.env` and credential stores are refused; templates and public keys remain readable. Only UTF-8 text without NUL bytes is accepted; use Bash or an MCP tool for other binaries.\n\nOutput is `<line-number>\\t<content>` per line. The trailing `<system>...</system>` block is status, not file content. Pure CRLF appears as LF and is preserved by Edit. Mixed or lone carriage returns appear as `\\r` and require exact `Edit.old_string` escapes.";
 	READ_DESCRIPTION = renderPrompt(read_default, {
 		MAX_LINES: MAX_LINES$1,
 		MAX_BYTES_KB: MAX_BYTES / 1024,
@@ -262799,7 +262799,7 @@ var init_turn = __esmMin((() => {
 											mode: todoMaintenanceMode,
 											workCallsSinceRefresh: this.agent.goalTodoPolicyState?.workCallsSinceRefresh ?? 0
 									});
-									const todoSystemPrompt = buildGoalTodoMaintenanceSystemPrompt(this.agent, turnSystemPrompt ?? this.agent.effectiveSystemPrompt, todoMaintenanceMode);
+									const todoSystemPrompt = buildGoalTodoMaintenanceSystemPrompt(this.agent, todoMaintenanceMode);
 									const todoMessages = buildGoalTodoMaintenanceMessages(this.agent, todoMaintenanceMode);
 									return {
 										llm: this.agent.llmForTurn("low", todoSystemPrompt),
@@ -423267,7 +423267,7 @@ function goalTodoMaintenanceMode(agent) {
 	};
 	return progress.refreshRequired || progress.workCallsSinceRefresh >= GOAL_TODO_REFRESH_WORK_CALL_LIMIT ? "refresh" : null;
 }
-function buildGoalTodoMaintenanceSystemPrompt(agent, basePrompt, mode) {
+function buildGoalTodoMaintenanceSystemPrompt(agent, mode) {
 	const todos = ideaTodos(agent);
 	const visibleTodoList = todos.length === 0 ? "The visible TodoList is empty." : ["Current visible TodoList:", ...todos.map((todo) => `- [${String(todo?.status ?? "pending")}] ${String(todo?.title ?? "").trim()}`)].join("\n");
 	const action = mode === "initial" ? "Create the visible task-specific TodoList now. Keep exactly one item in_progress and every later item pending." : "Update the visible TodoList now, before any more task work. Mark only evidenced steps done, keep the actual current step in_progress, and leave later steps pending. If the current title is stale, refine it to the concrete verified substep.";
@@ -423277,9 +423277,7 @@ ${action}
 Call TodoList as the first and only tool. Do not call a work, search, file, shell, web, messaging, or goal tool in this step. Do not merely describe the update.
 
 ${visibleTodoList}
-</todo-maintenance>
-
-${basePrompt}`;
+</todo-maintenance>`;
 }
 function buildGoalTodoMaintenanceMessages(agent, mode) {
 	const goal = agent.goal.getActiveGoal();
@@ -509237,7 +509235,7 @@ var SessionEventHandler = class {
 		if (matchedCall !== void 0 && matchedCall.name === "TodoList" && !event.isError) {
 			const rawTodos = matchedCall.args.todos;
 			if (Array.isArray(rawTodos)) {
-				const sanitized = rawTodos.filter((todo) => isTodoItemShape(todo)).map((t) => ({
+				const sanitized = rawTodos.filter((todo) => isTodoItemShape(todo)).filter((todo) => todo.status !== "done").map((t) => ({
 					title: t.title,
 					status: t.status
 				}));

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "blun-king-cli",
-  "version": "9.1.505",
+  "version": "9.1.507",
   "description": "BLUN CLI - your own AI agent with a Telegram channel. Get it done. With BLUN.",
   "license": "MIT",
   "bin": {