⬆️ chore: traer PRs upstream #177, #178, #151, #160

CHANGELOG.md

@@ -7,6 +7,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Added
+- **`/schedule` command**: Create, list, pause, resume, and remove scheduled jobs directly from Telegram. Auto-populates chat, directory, and user from context. Requires `ENABLE_SCHEDULER=true` (#150)
+
 ## [1.6.0] - 2026-03-30
 
 ### Added

CLAUDE.md

@@ -181,7 +181,7 @@ All datetimes use timezone-aware UTC: `datetime.now(UTC)` (not `datetime.utcnow(
 
 ### Agentic mode
 
-Agentic mode commands: `/start`, `/new`, `/status`, `/verbose`, `/repo`. If `ENABLE_PROJECT_THREADS=true`: `/sync_threads`. To add a new command:
+Agentic mode commands: `/start`, `/new`, `/status`, `/verbose`, `/repo`. If `ENABLE_PROJECT_THREADS=true`: `/sync_threads`. If `ENABLE_SCHEDULER=true`: `/schedule`. To add a new command:
 
 1. Add handler function in `src/bot/orchestrator.py`
 2. Register in `MessageOrchestrator._register_agentic_handlers()`

README.md

@@ -99,7 +99,7 @@ The bot supports two interaction modes:
 The default conversational mode. Just talk to Claude naturally -- no special commands required.
 
 **Commands:** `/start`, `/new`, `/status`, `/verbose`, `/repo`
-If `ENABLE_PROJECT_THREADS=true`: `/sync_threads`
+If `ENABLE_PROJECT_THREADS=true`: `/sync_threads` | If `ENABLE_SCHEDULER=true`: `/schedule`
 
 ```
 You: What files are in this project?
@@ -157,8 +157,8 @@ Use `/repo` to list cloned repos in your workspace, or `/repo <name>` to switch
 
 Set `AGENTIC_MODE=false` to enable the full 13-command terminal-like interface with directory navigation, inline keyboards, quick actions, git integration, and session export.
 
-**Commands:** `/start`, `/help`, `/new`, `/continue`, `/end`, `/status`, `/cd`, `/ls`, `/pwd`, `/projects`, `/export`, `/actions`, `/git`  
-If `ENABLE_PROJECT_THREADS=true`: `/sync_threads`
+**Commands:** `/start`, `/help`, `/new`, `/continue`, `/end`, `/status`, `/cd`, `/ls`, `/pwd`, `/projects`, `/export`, `/actions`, `/git`
+If `ENABLE_PROJECT_THREADS=true`: `/sync_threads` | If `ENABLE_SCHEDULER=true`: `/schedule`
 
 ```
 You: /cd my-web-app
@@ -176,7 +176,7 @@ Bot: [Run Tests] [Install Deps] [Format Code] [Run Linter]
 Beyond direct chat, the bot can respond to external triggers:
 
 - **Webhooks** -- Receive GitHub events (push, PR, issues) and route them through Claude for automated summaries or code review
-- **Scheduler** -- Run recurring Claude tasks on a cron schedule (e.g., daily code health checks)
+- **Scheduler** -- Run recurring Claude tasks on a cron schedule (e.g., daily code health checks). Manage jobs directly from Telegram with `/schedule`
 - **Notifications** -- Deliver agent responses to configured Telegram chats
 
 Enable with `ENABLE_API_SERVER=true` and `ENABLE_SCHEDULER=true`. See [docs/setup.md](docs/setup.md) for configuration.

docs/setup.md

@@ -173,7 +173,23 @@ ENABLE_SCHEDULER=true
 NOTIFICATION_CHAT_IDS=123456789  # Where to deliver results
 ```
 
-Jobs are managed programmatically and persist in the SQLite database.
+Jobs persist in the SQLite database and survive bot restarts. Manage them directly from Telegram with the `/schedule` command:
+
+```
+/schedule list                                      # List all jobs (active + paused)
+/schedule add <name> <min> <hour> <day> <month> <weekday> <prompt>
+/schedule remove <job_id>                           # Remove a job
+/schedule pause <job_id>                            # Pause without deleting
+/schedule resume <job_id>                           # Resume a paused job
+```
+
+Example -- create a job that runs every weekday at 9 AM:
+
+```
+/schedule add daily-report 0 9 * * 1-5 Summarize yesterday's git commits
+```
+
+The `chat_id`, `working_directory`, and `created_by` fields are auto-populated from your current Telegram context. Results are delivered to the chat where the job was created.
 
 ### Voice Message Transcription
 

src/bot/handlers/callback.py

@@ -12,6 +12,7 @@
 from ...security.audit import AuditLogger
 from ...security.validators import SecurityValidator
 from ..utils.html_format import escape_html
+from .command import _handle_model_selection
 
 logger = structlog.get_logger()
 
@@ -66,6 +67,8 @@ async def handle_callback_query(
             "conversation": handle_conversation_callback,
             "git": handle_git_callback,
             "export": handle_export_callback,
+            "model": lambda q, p, ctx: _handle_model_selection(q, f"model:{p}", ctx),
+            "effort": lambda q, p, ctx: _handle_model_selection(q, f"effort:{p}", ctx),
         }
 
         handler = handlers.get(action)

src/bot/handlers/command.py

@@ -7,7 +7,7 @@
 from typing import Optional
 
 import structlog
-from telegram import InlineKeyboardButton, InlineKeyboardMarkup, Update
+from telegram import CallbackQuery, InlineKeyboardButton, InlineKeyboardMarkup, Update
 from telegram.ext import ContextTypes
 
 from ...claude.facade import ClaudeIntegration
@@ -174,7 +174,8 @@ async def help_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> No
         "• <code>/status</code> - Show session and usage status\n"
         "• <code>/export</code> - Export session history\n"
         "• <code>/actions</code> - Show context-aware quick actions\n"
-        "• <code>/git</code> - Git repository information\n\n"
+        "• <code>/git</code> - Git repository information\n"
+        "• <code>/model [name]</code> - View or switch Claude model\n\n"
         "<b>Session Behavior:</b>\n"
         "• Sessions are automatically maintained per project directory\n"
         "• Switching directories with <code>/cd</code> resumes the session for that project\n"
@@ -1232,6 +1233,165 @@ async def git_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> Non
         logger.error("Error in git_command", error=str(e), user_id=user_id)
 
 
+# Short CLI aliases passed directly to the Claude CLI, which resolves them to
+# the current latest model of each family. No version numbers to maintain here.
+# See: https://docs.anthropic.com/en/docs/about-claude/models/overview
+_MODEL_FAMILIES = ["opus", "sonnet", "haiku"]
+
+# Effort levels per model family. Haiku has none; "max" is Opus-only.
+# Update here if a future model's effort support changes.
+_EFFORT_BY_MODEL = {
+    "opus": ["low", "medium", "high", "max"],
+    "sonnet": ["low", "medium", "high"],
+    "haiku": [],
+}
+
+
+def _current_model_label(context: ContextTypes.DEFAULT_TYPE) -> str:
+    """Return a human-friendly label for the active model + effort."""
+    override = context.user_data.get("model_override")  # "opus", "sonnet", "haiku", or None
+    effort = context.user_data.get("effort_override")
+    if not override:
+        settings = context.bot_data.get("settings")
+        server_model = getattr(settings, "claude_model", None) if settings else None
+        label = f"Default ({server_model or 'CLI default'})"
+    else:
+        label = override.capitalize()
+    return f"{label} | effort={effort}" if effort else label
+
+
+async def model_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
+    """Handle /model command - show model selection keyboard."""
+    current = _current_model_label(context)
+
+    keyboard = [
+        [
+            InlineKeyboardButton("Opus", callback_data="model:opus"),
+            InlineKeyboardButton("Sonnet", callback_data="model:sonnet"),
+            InlineKeyboardButton("Haiku", callback_data="model:haiku"),
+        ],
+        [InlineKeyboardButton("Reset to default", callback_data="model:default")],
+    ]
+
+    await update.message.reply_text(
+        f"🤖 <b>Current:</b> {escape_html(current)}\n\n"
+        "Choose a model:\n"
+        "<i>⚠️ Switching will start a new session.</i>",
+        parse_mode="HTML",
+        reply_markup=InlineKeyboardMarkup(keyboard),
+    )
+
+
+async def _handle_model_selection(
+    query: CallbackQuery,
+    data: str,
+    context: ContextTypes.DEFAULT_TYPE,
+) -> None:
+    """Shared logic for model/effort selection (used by both callback routes)."""
+    if data.startswith("model:"):
+        choice = data.split(":", 1)[1]
+
+        if choice == "default":
+            context.user_data.pop("model_override", None)
+            context.user_data.pop("effort_override", None)
+            # Note: if PR #165 merges first, change this to context.chat_data
+            context.user_data["force_new_session"] = True
+            await query.edit_message_text(
+                "🤖 Model and effort reset to server defaults.\n"
+                "<i>Next message starts a fresh session.</i>",
+                parse_mode="HTML",
+            )
+            logger.info("Model override cleared", user_id=query.from_user.id)
+            return
+
+        if choice not in _MODEL_FAMILIES:
+            await query.edit_message_text("Unknown model.")
+            return
+
+        # Store short CLI alias ("opus"/"sonnet"/"haiku") — the CLI resolves it
+        # to the current latest model, so no version numbers to maintain.
+        context.user_data["model_override"] = choice
+        # Clear stale effort when switching models
+        context.user_data.pop("effort_override", None)
+        # Force new session so the model change takes effect immediately
+        # Note: if PR #165 merges first, change this to context.chat_data
+        context.user_data["force_new_session"] = True
+
+        logger.info(
+            "Model override set",
+            user_id=query.from_user.id,
+            model=choice,
+        )
+
+        # Show effort level selection (if supported by this model)
+        effort_levels = _EFFORT_BY_MODEL.get(choice, [])
+        if not effort_levels:
+            # Model doesn't support effort (e.g. Haiku)
+            current = _current_model_label(context)
+            await query.edit_message_text(
+                f"🤖 <b>{escape_html(current)}</b> — ready.\n"
+                "<i>New session will start with your next message.</i>",
+                parse_mode="HTML",
+            )
+            return
+
+        rows = []
+        row = []
+        for level in effort_levels:
+            row.append(
+                InlineKeyboardButton(level.capitalize(), callback_data=f"effort:{level}")
+            )
+            if len(row) == 2:
+                rows.append(row)
+                row = []
+        if row:
+            rows.append(row)
+        rows.append(
+            [InlineKeyboardButton("Skip (keep current)", callback_data="effort:skip")]
+        )
+
+        await query.edit_message_text(
+            f"🤖 Model set to <b>{escape_html(choice.capitalize())}</b>.\n\n"
+            "Choose effort level:",
+            parse_mode="HTML",
+            reply_markup=InlineKeyboardMarkup(rows),
+        )
+
+    elif data.startswith("effort:"):
+        level = data.split(":", 1)[1]
+
+        if level == "skip":
+            current = _current_model_label(context)
+            await query.edit_message_text(
+                f"🤖 <b>{escape_html(current)}</b> — ready.\n"
+                "<i>New session will start with your next message.</i>",
+                parse_mode="HTML",
+            )
+            return
+
+        all_effort_levels = {"low", "medium", "high", "max"}
+        if level in all_effort_levels:
+            context.user_data["effort_override"] = level
+            current = _current_model_label(context)
+            await query.edit_message_text(
+                f"🤖 <b>{escape_html(current)}</b> — ready.\n"
+                "<i>New session will start with your next message.</i>",
+                parse_mode="HTML",
+            )
+            logger.info(
+                "Effort override set",
+                user_id=query.from_user.id,
+                effort=level,
+            )
+
+
+async def model_callback(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
+    """Handle model and effort selection callbacks (agentic mode route)."""
+    query = update.callback_query
+    await query.answer()
+    await _handle_model_selection(query, query.data, context)
+
+
 async def restart_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
     """Handle /restart command - gracefully restart the bot process.
 

src/bot/handlers/message.py

@@ -393,6 +393,8 @@ async def stream_handler(update_obj):
                 session_id=session_id,
                 on_stream=stream_handler,
                 force_new=force_new,
+                model_override=context.user_data.get("model_override"),
+                effort_override=context.user_data.get("effort_override"),
             )
 
             # New session created successfully — clear the one-shot flag
@@ -818,6 +820,8 @@ async def handle_document(update: Update, context: ContextTypes.DEFAULT_TYPE) ->
                 working_directory=current_dir,
                 user_id=user_id,
                 session_id=session_id,
+                model_override=context.user_data.get("model_override"),
+                effort_override=context.user_data.get("effort_override"),
             )
 
             # Update session ID
@@ -945,6 +949,8 @@ async def handle_photo(update: Update, context: ContextTypes.DEFAULT_TYPE) -> No
                     working_directory=current_dir,
                     user_id=user_id,
                     session_id=session_id,
+                    model_override=context.user_data.get("model_override"),
+                    effort_override=context.user_data.get("effort_override"),
                 )
 
                 # Update session ID
@@ -1073,6 +1079,8 @@ async def handle_voice(update: Update, context: ContextTypes.DEFAULT_TYPE) -> No
                 working_directory=current_dir,
                 user_id=user_id,
                 session_id=session_id,
+                model_override=context.user_data.get("model_override"),
+                effort_override=context.user_data.get("effort_override"),
             )
 
             context.user_data["claude_session_id"] = claude_response.session_id