agentwire-dev 1.28.1__tar.gz → 1.29.1__tar.gz

{agentwire_dev-1.28.1 → agentwire_dev-1.29.1}/CHANGELOG.md

@@ -9,6 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Removed
 
+- **Workflow engine** — entire `agentwire/workflows/` package (DAG runner, definitions, storage, pi runner, CLI), the `agentwire workflow` CLI subcommand, 5 MCP tools (`workflow_list/validate/run/history/show`), 3 portal endpoints (`/api/workflows/list|runs|runs/{id}`), the Workflows sidebar section + WorkflowWindow + ~230 lines of CSS, the scheduler's `workflow:` dispatch branch (`SchedulerTask.workflow`/`inputs` fields, `_dispatch_workflow_task`, `_render_workflow_inputs`, `_parse_workflow_summary`, `_workflow_failure_state`, validation block), 9 workflow test files plus the workflow-specific cases in `test_scheduler.py`, the `agentwire-workflows` skill, `docs/wiki/scheduling/workflows.md`, and ~120 dirs of historical runs under `~/.agentwire/workflows/`. The user's actual recurring automations all live as scheduler ensure tasks (`task:`); the engine was unused in practice and the empty sidebar surface was just noise. Scheduler is now ensure-only.
 - **Inbound channel surface** — entire Telegram bridge (`agentwire/bridges/telegram.py`), Discord channel (`agentwire/channels/discord.py`), Slack channel (`agentwire/channels/slack.py`), Twilio SMS (`agentwire/channels/sms.py`), Webhook (`agentwire/channels/webhook.py`), outbound-Telegram (`agentwire/channels/telegram.py`), channel scaffolding template (`agentwire/channels/_template.py`), all `agentwire {telegram,discord,slack,sms,webhook}` CLI command groups, the `agentwire reply` command, MCP tools `discord_status` / `slack_status` / `sms_send` / `webhook_send`, the portal sidebar "socials" section, the `ServiceChannel` / `MessageQueueManager` / `QueuedMessage` / `compose_session_config` / `inject_instructions` / session-helper plumbing in `channels/base.py`, the TTS/STT primitives on the channel base class, the auto-injected `discord-dm` / `slack-dm` / `channel-admin` roles, the `OutputConfig.notify` task field plus its `_handle_task_notification` dispatcher (voice / alert / webhook / email / command), the doctor's Telegram bridge check, and the `aiogram` optional dep. Wire surface is now outbound-only: **email** (Resend) and **quo** (OpenPhone SMS) remain. Inbound user input flows through the portal. Security review traced the cut: Telegram bridge was secure (refuse-to-start on empty allowlist + silent default-deny), Discord/Slack fail-open on empty allowlist + unauthenticated channel `@mention`; rather than fix the gaps, the whole inbound surface comes out to keep the wire posture outbound-only.
 - **Anthropic SDK surface** — entire `agentwire/sdk/` package, `agentwire/repl/` (Textual REPL), `agentwire/workflows/runners/{anthropic,human_gate}.py`, `sdk-bypass` / `sdk-prompted` / `sdk-restricted` session types, `sdk-watch` portal window + sidebar section, `--runner` workflow CLI override, and the `claude-agent-sdk` dependency. Anthropic moved Agent SDK + `claude -p` usage out of the monthly subscription tier on 2026-05-13 (now API-billed); the in-tree consumers were too expensive to keep running. The `pi` workflow runner is now the only runner. For Anthropic-quality scheduled work, use a `claude-bypass` tmux session via `.agentwire.yml` + scheduler. Mission [#184](https://github.com/dotdevdotdev/agentwire-dev/issues/184). Findings captured in `~/.agentwire/wiki/` (anthropic-agent-sdk, anthropic-runner-findings, textual-repl-retrospective, workflow-runner-choice, claude-code-print-mode).
 

{agentwire_dev-1.28.1 → agentwire_dev-1.29.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentwire-dev
-Version: 1.28.1
+Version: 1.29.1
 Summary: Multi-session voice web interface for AI coding agents
 Project-URL: Homepage, https://agentwire.dev
 Project-URL: Repository, https://github.com/dotdevdotdev/agentwire-dev

agentwire_dev-1.29.1/SECURITY.md

@@ -0,0 +1,59 @@
+# Security Policy
+
+## Reporting a Vulnerability
+
+If you discover a security vulnerability in AgentWire, please report it privately.
+
+**Do NOT open a public GitHub issue for security vulnerabilities.**
+
+### How to Report
+
+Email: security@agentwire.dev
+
+Include:
+- Description of the vulnerability
+- Steps to reproduce
+- Potential impact
+- Any suggested fixes (optional)
+
+### What to Expect
+
+- **Acknowledgment:** Within 48 hours
+- **Initial Assessment:** Within 1 week
+- **Resolution Timeline:** Depends on severity, typically 30-90 days
+
+### Scope
+
+This security policy applies to:
+- The AgentWire CLI (`agentwire` command)
+- The AgentWire portal (web interface)
+- Official AgentWire packages on PyPI
+
+### Out of Scope
+
+- Third-party dependencies (report to their maintainers)
+- Self-hosted TTS/STT servers
+- User misconfiguration
+
+## Security Features
+
+AgentWire includes built-in security features:
+
+- **Damage Control Hooks:** Block 300+ dangerous command patterns
+- **Path Protection:** Prevent access to sensitive files (.env, SSH keys, credentials)
+- **Audit Logging:** All blocked operations are logged
+
+See `docs/wiki/internals/damage-control.md` for details.
+
+## Trust Model
+
+The portal (HTTPS server bound to `0.0.0.0:8765` by default) **has no built-in authentication or authorization on its API endpoints**. This is by design — agentwire is built for a local-network trust perimeter, typically running on the same machine as the operator's browser or behind a Cloudflare Tunnel + Zero Trust gate (see `docs/wiki/deployment/remote-access.md`).
+
+What this means in practice:
+
+- **Anyone who can reach the portal port can drive it.** All `/api/*` endpoints (scheduler control, missions dispatch, project deletion, artifact upload, desktop window control) execute without auth.
+- **Do not expose the portal directly to the public internet.** Use either firewall rules limiting access to trusted IPs, or front it with an auth gateway. Cloudflare Tunnel + Zero Trust is the recommended pattern; details in the deployment docs.
+- **Project deletion via `/api/projects/delete`** validates the path is absolute, contains no `..`, contains no shell metacharacters, and is not in a protected list. Local execution uses argv form (no shell); SSH execution uses `shlex.quote` per argument. These mitigations don't substitute for perimeter security — they reduce blast radius if the perimeter fails.
+- **CSRF / Origin checks are not enforced** on state-changing POSTs. A browser inside the trust perimeter that loads attacker-controlled content could be coerced into making requests to the portal. If your portal is reachable from a browser on a less-trusted network, add an origin check or fence it behind an auth proxy.
+
+If you need authentication, the recommended path is **Cloudflare Tunnel + Zero Trust** rather than adding auth in-process: identity, MFA, audit, and revocation are all handled upstream and survive process restarts.

{agentwire_dev-1.28.1 → agentwire_dev-1.29.1}/agentwire/__init__.py

@@ -1,3 +1,3 @@
 """AgentWire - Multi-session voice web interface for AI coding agents."""
 
-__version__ = "1.28.1"
+__version__ = "1.29.1"

{agentwire_dev-1.28.1 → agentwire_dev-1.29.1}/agentwire/__main__.py

@@ -2,7 +2,6 @@
 
 import argparse
 import base64
-from dataclasses import dataclass, field
 import datetime
 import importlib.resources
 import json
@@ -17,6 +16,7 @@ import tempfile
 import time
 import urllib.error
 import urllib.request
+from dataclasses import dataclass, field
 from pathlib import Path
 
 from dotenv import load_dotenv
@@ -101,6 +101,17 @@ def _build_tmux_env_flags_shell(env: dict[str, str]) -> str:
     return " ".join(parts) + " "
 
 
+def _set_session_name_env(agent: "AgentCommand", session_name: str) -> None:
+    """Stamp ``AGENTWIRE_SESSION_NAME`` onto an ``AgentCommand.env``.
+
+    Every session created via ``cmd_new`` / ``cmd_spawn`` / ``cmd_recreate``
+    / ``cmd_fork`` / scheduler-spawn paths gets this so downstream tooling
+    (notably the mission-worker damage-control rules in ``safety/_core.py``)
+    can identify which agentwire session the running tool is part of.
+    """
+    agent.env["AGENTWIRE_SESSION_NAME"] = session_name
+
+
 def inject_session_env(session: str, env: dict[str, str], remote_host: str | None = None) -> None:
     """Set env vars on an existing tmux session for FUTURE shells in that session.
 
@@ -1386,9 +1397,10 @@ def cmd_tts_restart(args) -> int:
     Used by CLI when venv_mismatch occurs during TTS request.
     Supports both local and remote TTS servers.
     """
-    from .network import NetworkContext
     import time
 
+    from .network import NetworkContext
+
     ctx = NetworkContext.from_config()
     session_name = get_tts_session_name()
 
@@ -2907,40 +2919,14 @@ def cmd_send(args) -> int:
             print(f"Session '{session}' not found", file=sys.stderr)
         return 1
 
-    # Send the prompt — use load-buffer for anything non-trivial to avoid PTY flooding
-    use_buffer = len(prompt) > 10 or "\n" in prompt
-
-    if use_buffer:
-        with tempfile.NamedTemporaryFile(mode='w', suffix='.txt', delete=False) as f:
-            f.write(prompt)
-            temp_path = f.name
-        try:
-            subprocess.run(["tmux", "load-buffer", temp_path], check=True)
-            subprocess.run(["tmux", "paste-buffer", "-t", session], check=True)
-        finally:
-            os.unlink(temp_path)
-    else:
-        subprocess.run(
-            ["tmux", "send-keys", "-t", session, "-l", prompt],
-            check=True
-        )
-
-    # Wait for text to be displayed before pressing Enter
-    time.sleep(0.5)
-
-    subprocess.run(
-        ["tmux", "send-keys", "-t", session, "Enter"],
-        check=True
-    )
-
-    # For multi-line text, Claude Code shows "[Pasted text...]" and waits for Enter
-    # Send another Enter after a short delay to confirm the paste
-    if "\n" in prompt or len(prompt) > 200:
-        time.sleep(0.5)
-        subprocess.run(
-            ["tmux", "send-keys", "-t", session, "Enter"],
-            check=True
-        )
+    # Delegate paste + Enter handling to the shared pane_manager helper so
+    # send-to-session and send-to-pane stay in lockstep. The helper handles
+    # buffer-vs-send-keys, settle delay, and the bracketed-paste double-Enter
+    # in one place.
+    try:
+        pane_manager.send_to_target(session, prompt, enter=True)
+    except RuntimeError as e:
+        return _output_result(False, json_mode, str(e))
 
     if json_mode:
         print(json.dumps({"success": True, "session": session_full, "machine": None, "message": "Prompt sent"}))
@@ -3372,6 +3358,40 @@ def cmd_new(args) -> int:
         # Simple session: ~/projects/project/
         session_path = projects_dir / project
 
+    # Safety: refuse to attach a new session to a path that is already the working
+    # directory of another active session. Two agents sharing the same working tree
+    # is the dangerous footgun (one's dirty state visible to the other, branches
+    # mixing). Worktree sessions (project/branch) get unique paths and won't trip
+    # this. --force overrides.
+    if not args.force:
+        target = str(session_path.resolve()) if session_path.exists() else str(session_path)
+        panes_result = subprocess.run(
+            ["tmux", "list-panes", "-a", "-F", "#{session_name}\t#{pane_current_path}"],
+            capture_output=True, text=True,
+        )
+        if panes_result.returncode == 0:
+            conflicting: set[str] = set()
+            for line in panes_result.stdout.splitlines():
+                if "\t" not in line:
+                    continue
+                sess, p = line.split("\t", 1)
+                if sess == session_name:
+                    continue
+                try:
+                    if str(Path(p).resolve()) == target:
+                        conflicting.add(sess)
+                except (OSError, RuntimeError):
+                    continue
+            if conflicting:
+                others = ", ".join(sorted(conflicting))
+                hint = (
+                    f"Refusing to attach session '{session_name}' to {session_path}: "
+                    f"already the working directory of active session(s): {others}. "
+                    "Use a 'project/branch' name to create an isolated worktree, "
+                    "pick a different path, or pass --force to override."
+                )
+                return _output_result(False, json_mode, hint)
+
     if not session_path.exists():
         if args.force or path:
             # Auto-create directory with -f flag or when custom path explicitly provided
@@ -3431,6 +3451,7 @@ def cmd_new(args) -> int:
     model_override = getattr(args, 'model', None)
     agent = build_agent_command(session_type, roles if roles else None, model=model_override)
     agent.env.update(parse_env_args(getattr(args, 'env', None)))
+    _set_session_name_env(agent, session_name)
 
     agent_cmd = agent.command
 
@@ -6291,9 +6312,10 @@ def cmd_voiceclone_list(args) -> int:
     """List available voices."""
     json_mode = getattr(args, 'json', False)
 
-    from .voiceclone import is_runpod_backend, list_voices_runpod, get_tts_url
     import requests
 
+    from .voiceclone import get_tts_url, is_runpod_backend, list_voices_runpod
+
     if is_runpod_backend():
         success, result = list_voices_runpod()
         if success:
@@ -7754,22 +7776,16 @@ def cmd_ensure(args) -> int:
     10. Handle retries on failure
     """
     from .completion import (
-        CompletionTimeout,
-        generate_summary_filename,
         get_summary_prompt,
-        status_to_exit_code,
     )
     from .locking import LockConflict, LockTimeout, session_lock
     from .tasks import (
-        PreCommandError,
         TaskNotFound,
         TaskValidationError,
         load_task,
-        run_post_command,
-        run_pre_command,
         validate_task,
     )
-    from .templating import TemplateContext, TemplateError, expand_all, preview_template
+    from .templating import TemplateContext, preview_template
 
     session_name = args.session
     task_name = args.task
@@ -8109,7 +8125,7 @@ def _run_ensure_task(args, session, task, ctx, shell, project_path, json_mode) -
                         capture_output=True, text=True,
                     )
                     if fork_result.returncode != 0 and not json_mode:
-                        print(f"Warning: context fork failed, starting fresh session")
+                        print("Warning: context fork failed, starting fresh session")
                 elif not json_mode:
                     print(f"Warning: starting_session '{task.starting_session}' not found, starting fresh")
 
@@ -8813,10 +8829,9 @@ def cmd_scheduler_board(args) -> int:
 
 def cmd_scheduler_run(args) -> int:
     """Force-run a specific task now."""
-    from .scheduler import _render_workflow_inputs, dispatch_task, load_board, save_board
+    from .scheduler import dispatch_task, load_board, save_board
 
     json_mode = getattr(args, 'json', False)
-    dry_run = getattr(args, 'dry_run', False)
     name = args.name
 
     try:
@@ -8830,40 +8845,6 @@ def cmd_scheduler_run(args) -> int:
             f"Task '{name}' not found in board. Available: {', '.join(board.tasks.keys())}",
         )
 
-    task = board.tasks[name]
-
-    if dry_run:
-        if not task.workflow:
-            return _output_result(
-                False, json_mode,
-                f"--dry-run only applies to workflow tasks; '{name}' runs via ensure.",
-            )
-        from agentwire.workflows.definitions import resolve_workflow
-        from agentwire.workflows.runner import run_workflow
-        try:
-            wf = resolve_workflow(task.workflow)
-        except Exception as e:
-            return _output_result(False, json_mode, f"Could not load workflow '{task.workflow}': {e}")
-        rendered_inputs = _render_workflow_inputs(task.inputs, task)
-        run = run_workflow(wf, runs_dir=None, inputs=rendered_inputs, dry_run=True)
-        if json_mode:
-            _output_json({
-                "success": True,
-                "task": name,
-                "dry_run": True,
-                "workflow": wf.name,
-                "inputs": rendered_inputs,
-                "nodes": [r.node_id for r in run.node_results],
-                "status": run.status,
-            })
-            return 0
-        print(f"Dry-run: {name} → workflow '{wf.name}'")
-        if rendered_inputs:
-            print(f"  inputs: {rendered_inputs}")
-        for r in run.node_results:
-            print(f"  - {r.node_id}: {r.final_text}")
-        return 0
-
     if not json_mode:
         print(f"Running: {name}")
 
@@ -8946,7 +8927,6 @@ def cmd_scheduler_history(args) -> int:
                 "last_status": state.last_status,
                 "last_duration": state.last_duration,
                 "run_count": state.run_count,
-                "workflow": (task.workflow if task else "") or None,
             })
         _output_json({"success": True, "history": history})
         return 0
@@ -8972,8 +8952,8 @@ def cmd_scheduler_history(args) -> int:
 
 def cmd_scheduler_report(args) -> int:
     """Generate a morning report HTML artifact of recent task runs."""
-    import re as _re
     from html import escape as html_escape
+
     from .scheduler import _parse_duration, format_interval, load_board, read_events
 
     json_mode = getattr(args, 'json', False)
@@ -9045,38 +9025,11 @@ def cmd_scheduler_report(args) -> int:
         color = colors.get(status, "#78909c")
         return f'<span style="background:{color};color:#fff;padding:2px 8px;border-radius:12px;font-size:0.85em">{status}</span>'
 
-    def node_badges(nodes: list) -> str:
-        if not nodes:
-            return ""
-        node_colors = {
-            "success": "#00c853",
-            "failure": "#ff5252",
-            "timeout": "#ff7043",
-            "skipped": "#78909c",
-        }
-        pieces = []
-        for n in nodes:
-            nid = html_escape(str(n.get("id", "?")))
-            nstatus = str(n.get("status", "?"))
-            color = node_colors.get(nstatus, "#556")
-            pieces.append(
-                f'<span style="background:{color};color:#fff;padding:1px 6px;border-radius:8px;font-size:0.75em;margin-right:3px">{nid}</span>'
-            )
-        return "".join(pieces)
-
     rows_html = ""
     for r in runs:
         duration_str = format_interval(r["duration"]) if r["duration"] else "-"
         pr_link = f'<a href="{r["pr_url"]}" target="_blank" style="color:#00d4ff">{r["pr_url"][:40]}...</a>' if r.get("pr_url") else "-"
-        branch_col = ""
-        if r.get("workflow"):
-            wf_label = html_escape(r["workflow"])
-            run_id = html_escape(r.get("run_id", ""))
-            badges = node_badges(r["nodes"])
-            run_hint = f' <span style="color:#556;font-size:0.75em">({run_id[:20]}…)</span>' if run_id else ""
-            branch_col = f'<code style="font-size:0.85em">workflow:{wf_label}</code>{run_hint}<div style="margin-top:4px">{badges}</div>'
-        else:
-            branch_col = f'<code style="font-size:0.85em">{r.get("work_branch") or "-"}</code>'
+        branch_col = f'<code style="font-size:0.85em">{r.get("work_branch") or "-"}</code>'
         summary_text = r["summary"][:120] if r["summary"] else "-"
         rows_html += f"""
         <tr>
@@ -9536,7 +9489,7 @@ def cmd_overnight_prepare(args) -> int:
             "status": item.status,
         })
     else:
-        print(f"Queued for overnight execution:")
+        print("Queued for overnight execution:")
         print(f"  ID: {item.id}")
         print(f"  Session: {item.session}")
         print(f"  Branch: {item.work_branch}")
@@ -9548,7 +9501,7 @@ def cmd_overnight_prepare(args) -> int:
 
 def cmd_overnight_list(args) -> int:
     """List overnight queue items."""
-    from .overnight import load_queue, load_done
+    from .overnight import load_done, load_queue
 
     json_mode = getattr(args, "json", False)
     show_all = getattr(args, "all", False)
@@ -9587,7 +9540,7 @@ def cmd_overnight_list(args) -> int:
 
 def cmd_overnight_status(args) -> int:
     """Show overnight orchestrator state and queue summary."""
-    from .overnight import load_queue, read_live_state, in_overnight_window
+    from .overnight import in_overnight_window, load_queue, read_live_state
 
     json_mode = getattr(args, "json", False)
 
@@ -9633,7 +9586,7 @@ def cmd_overnight_status(args) -> int:
 
 def cmd_overnight_cancel(args) -> int:
     """Cancel a queued overnight item."""
-    from .overnight import load_item, save_item, delete_item
+    from .overnight import delete_item, load_item
 
     item_id = args.id
     json_mode = getattr(args, "json", False)
@@ -10209,70 +10162,6 @@ def main() -> int:
     )
     dev_parser.set_defaults(func=cmd_dev)
 
-    # === workflow command group ===
-    from agentwire.workflows.cli import (
-        cmd_workflow_history,
-        cmd_workflow_list,
-        cmd_workflow_run,
-        cmd_workflow_show,
-        cmd_workflow_validate,
-    )
-
-    workflow_parser = subparsers.add_parser("workflow", help="Pi workflow engine")
-    workflow_subparsers = workflow_parser.add_subparsers(dest="workflow_command")
-
-    wf_list = workflow_subparsers.add_parser("list", help="List discoverable workflows")
-    wf_list.add_argument("--json", action="store_true", help="Output as JSON")
-    wf_list.set_defaults(func=cmd_workflow_list)
-
-    wf_validate = workflow_subparsers.add_parser(
-        "validate", help="Validate a workflow YAML without running it"
-    )
-    wf_validate.add_argument("workflow", help="Workflow name or path to YAML")
-    wf_validate.set_defaults(func=cmd_workflow_validate)
-
-    wf_run = workflow_subparsers.add_parser("run", help="Execute a workflow")
-    wf_run.add_argument("workflow", help="Workflow name or path to YAML")
-    wf_run.add_argument(
-        "--input", action="append", metavar="KEY=VALUE",
-        help="Workflow input (repeatable). Overrides --input-file."
-    )
-    wf_run.add_argument(
-        "--input-file", metavar="PATH",
-        help="JSON file with inputs (object mapping name → value)"
-    )
-    wf_run.add_argument("--dry-run", action="store_true", help="Print plan without running")
-    wf_run.add_argument("--verbose", "-v", action="store_true", help="Verbose output")
-    wf_run.add_argument("--json", action="store_true", help="Output as JSON")
-    wf_run.set_defaults(func=cmd_workflow_run)
-
-    wf_history = workflow_subparsers.add_parser(
-        "history", help="List past workflow runs"
-    )
-    wf_history.add_argument(
-        "--workflow", metavar="NAME", help="Filter by workflow name"
-    )
-    wf_history.add_argument(
-        "--limit", type=int, default=20, help="Max runs to show (default: 20)"
-    )
-    wf_history.add_argument("--json", action="store_true", help="Output as JSON")
-    wf_history.set_defaults(func=cmd_workflow_history)
-
-    wf_show = workflow_subparsers.add_parser(
-        "show", help="Inspect a past workflow run"
-    )
-    wf_show.add_argument("run_id", help="Run ID (from `workflow history`)")
-    wf_show.add_argument(
-        "--events", action="store_true",
-        help="Dump raw event JSONL for all nodes",
-    )
-    wf_show.add_argument(
-        "--node", metavar="ID",
-        help="Filter events to one node id (implies --events)",
-    )
-    wf_show.add_argument("--json", action="store_true", help="Output as JSON")
-    wf_show.set_defaults(func=cmd_workflow_show)
-
     # === listen command group ===
     listen_parser = subparsers.add_parser("listen", help="Voice input recording")
     listen_parser.add_argument(
@@ -10738,8 +10627,6 @@ def main() -> int:
     sched_run = scheduler_subparsers.add_parser("run", help="Force-run a task now")
     sched_run.add_argument("name", help="Task name from board")
     sched_run.add_argument("--json", action="store_true", help="Output JSON")
-    sched_run.add_argument("--dry-run", action="store_true",
-                           help="For workflow tasks: print the execution plan without running")
     sched_run.set_defaults(func=cmd_scheduler_run)
 
     # scheduler enable <name>
@@ -10782,6 +10669,95 @@ def main() -> int:
     sched_report.add_argument("--json", action="store_true", help="Output JSON")
     sched_report.set_defaults(func=cmd_scheduler_report)
 
+    # === mission command group ===
+    from .missions import cli as mission_cli
+
+    mission_parser = subparsers.add_parser(
+        "mission",
+        help="Manage agent-driven missions (issue→branch→draft PR cycles)",
+        description=(
+            "First-class auto-dispatcher: stateless orchestrators tick the GitHub "
+            "issue board, spawn worker sessions in isolated worktrees, route PR "
+            "feedback back, and gc when the PR closes. See docs/MISSIONS.md."
+        ),
+    )
+    mission_subparsers = mission_parser.add_subparsers(dest="mission_command")
+
+    # mission list
+    m_list = mission_subparsers.add_parser("list", help="List active workers + eligible issues")
+    m_list.add_argument("--json", action="store_true", help="Output JSON")
+    m_list.set_defaults(func=mission_cli.cmd_mission_list)
+
+    # mission show <N>
+    m_show = mission_subparsers.add_parser("show", help="Show one issue + dispatch + PR state")
+    m_show.add_argument("number", type=int, help="GitHub issue number")
+    m_show.add_argument("--repo", required=True, help="Repo short name (from missions config)")
+    m_show.add_argument("--json", action="store_true", help="Output JSON")
+    m_show.set_defaults(func=mission_cli.cmd_mission_show)
+
+    # mission status
+    m_status = mission_subparsers.add_parser("status", help="Per-repo summary of active + eligible")
+    m_status.add_argument("--json", action="store_true", help="Output JSON")
+    m_status.set_defaults(func=mission_cli.cmd_mission_status)
+
+    # mission spawn <N>
+    m_spawn = mission_subparsers.add_parser(
+        "spawn", help="Force-dispatch an issue, bypassing eligibility"
+    )
+    m_spawn.add_argument("number", type=int, help="GitHub issue number")
+    m_spawn.add_argument("--repo", required=True, help="Repo short name")
+    m_spawn.add_argument("--json", action="store_true", help="Output JSON")
+    m_spawn.set_defaults(func=mission_cli.cmd_mission_spawn)
+
+    # mission stall <N>
+    m_stall = mission_subparsers.add_parser("stall", help="Stop dispatch for an issue with a reason")
+    m_stall.add_argument("number", type=int, help="GitHub issue number")
+    m_stall.add_argument("--repo", required=True, help="Repo short name")
+    m_stall.add_argument("--reason", required=True, help="Reason (posted as comment)")
+    m_stall.add_argument("--json", action="store_true", help="Output JSON")
+    m_stall.set_defaults(func=mission_cli.cmd_mission_stall)
+
+    # mission resume <N>
+    m_resume = mission_subparsers.add_parser("resume", help="Re-enable an issue for dispatch")
+    m_resume.add_argument("number", type=int, help="GitHub issue number")
+    m_resume.add_argument("--repo", required=True, help="Repo short name")
+    m_resume.add_argument("--json", action="store_true", help="Output JSON")
+    m_resume.set_defaults(func=mission_cli.cmd_mission_resume)
+
+    # mission kill <N>
+    m_kill = mission_subparsers.add_parser(
+        "kill", help="Kill worker session + worktree (does NOT close the PR)"
+    )
+    m_kill.add_argument("number", type=int, help="GitHub issue number")
+    m_kill.add_argument("--repo", required=True, help="Repo short name")
+    m_kill.add_argument("--json", action="store_true", help="Output JSON")
+    m_kill.set_defaults(func=mission_cli.cmd_mission_kill)
+
+    # mission gc
+    m_gc = mission_subparsers.add_parser("gc", help="Reap sessions whose PR is merged/closed")
+    m_gc.add_argument("--json", action="store_true", help="Output JSON")
+    m_gc.set_defaults(func=mission_cli.cmd_mission_gc)
+
+    # mission tick
+    m_tick = mission_subparsers.add_parser("tick", help="Run one dispatcher tick now")
+    m_tick.add_argument("--json", action="store_true", help="Output JSON")
+    m_tick.set_defaults(func=mission_cli.cmd_mission_tick)
+
+    # mission route-feedback
+    m_route = mission_subparsers.add_parser(
+        "route-feedback", help="Run one PR-feedback router tick now"
+    )
+    m_route.add_argument("--json", action="store_true", help="Output JSON")
+    m_route.set_defaults(func=mission_cli.cmd_mission_route_feedback)
+
+    # mission init <repo>
+    m_init = mission_subparsers.add_parser(
+        "init", help="Create the agent-ready label on a repo (idempotent)"
+    )
+    m_init.add_argument("repo", help="Repo short name (from config) or owner/repo form")
+    m_init.add_argument("--json", action="store_true", help="Output JSON")
+    m_init.set_defaults(func=mission_cli.cmd_mission_init)
+
     # === overnight command group ===
     overnight_parser = subparsers.add_parser(
         "overnight",
@@ -10918,6 +10894,10 @@ def main() -> int:
         scheduler_parser.print_help()
         return 0
 
+    if args.command == "mission" and getattr(args, "mission_command", None) is None:
+        mission_parser.print_help()
+        return 0
+
     if args.command == "overnight" and getattr(args, "overnight_command", None) is None:
         overnight_parser.print_help()
         return 0