Reference
CLI reference
On this page
Every public runner command and option below comes from the CLI shipped with this guide. For a walkthrough, see Drive Runner from agents.
- Command overview
- Agent guide
- runner
- runner daemon
- runner daemon status
- runner daemon stop
- runner status
- runner project
- runner project list
- runner project show
- runner project create
- runner project rename
- runner project delete
- runner role
- runner role list
- runner role show
- runner role create
- runner role update
- runner role delete
- runner crew
- runner crew list
- runner crew show
- runner crew create
- runner crew update
- runner crew delete
- runner crew add
- runner crew set
- runner crew remove
- runner crew lead
- runner crew order
- runner mission
- runner mission list
- runner mission show
- runner mission start
- runner mission add-role
- runner mission remove-role
- runner mission stop
- runner mission resume
- runner mission archive
- runner mission unarchive
- runner mission rename
- runner mission pin
- runner mission unpin
- runner mission move
- runner mission feed
- runner mission answer
- runner chat
- runner chat start
- runner session
- runner session list
- runner session show
- runner session stop
- runner session archive
- runner session resume
- runner session restart
- runner msg
- runner msg post
- runner msg read
- runner signal
- runner ask
- runner call
- runner help
Command overview
runner — operate Runner from a shell or a mission session
USAGE
runner status
runner daemon status|stop
runner project list|show|create|rename|delete
runner role list|show|create|update|delete
runner crew list|show|create|update|delete|add|set|remove|lead|order
runner mission list|show|start|stop|resume|archive|unarchive|rename|pin|unpin|move|feed|answer
runner chat start
runner session list|show|stop|archive|resume|restart
runner msg post|read
runner signal <type>
runner ask <question> | runner ask --human <prompt> --choices <a,b,...>
runner call <tool> [<json>]
OUTPUT
--json print the JSON result
-q print only result ids
CONTEXT
Inside a mission, msg post/read, signal, and ask use the event log directly.
msg read optionally queries the existing daemon for a read cursor; offline reads still work.
Outside, mission-scoped writes require --mission; --as names a roster handle.
RESERVED FOR #562
spawn, ps, wait, stop <handle>, done
Agent guide
runner — version-matched guide for agents
DISCOVER
runner status
runner project --help
runner role --help
runner crew --help
runner mission --help
runner chat --help
runner session --help
runner daemon --help
runner msg --help
runner signal --help
runner ask --help
runner call --help
runner help
OUTPUT AND EXITS
Add --json for structured output. Add -q when only the changed object's id is needed.
0 success; 1 Runner refused or watch failed; 2 usage or reference error; 3 Runner is not running;
5 a sandbox kept this command from reaching Runner.
On exit 3, ask the user to open Runner. On exit 5, nothing is known about Runner yet:
run the same command again outside the sandbox (in Codex, request escalated permissions
for it) before telling the user anything.
Use --help instead of guessing a command.
REFERENCES
Roles use their unique handle. Crews and projects accept an id or exact name.
Missions and direct-chat sessions accept an id or unique id prefix.
A mission or chat started inside a project's directory belongs to it; --project overrides.
COMMON FLOW
mission=$(runner mission start --title "Small task" -q)
runner mission add-role "$mission" <role> --task-file - < task.md
# Or start from a reusable crew:
runner crew list --json
mission=$(runner mission start --crew <crew> --goal-file - -q < brief.md)
runner mission feed "$mission" --follow --json
runner mission add-role "$mission" <role> --task-file - < task.md
runner mission remove-role "$mission" <handle>
runner mission show "$mission" --json
runner mission answer "$mission" <question_id_from_mission_show> <choice>
# Use the HANDLE from the LEAD row in mission show, or any member when there is no lead.
runner msg post --mission "$mission" --to <lead_handle> "message"
runner mission stop "$mission"
runner mission archive "$mission"
MESSAGING INSIDE A MISSION
Run runner msg read when an [inbox] notification arrives: only new inbox messages print,
followed by a short digest of others' directed traffic. Use runner msg read --all for all
crew messages since joining after a resume or when earlier messages are no longer in context.
Default reads already track unread messages: do not add --since or --from to consume your inbox.
--all acknowledges only your inbox. --since and --from are inspection filters and never mark
messages read, even with --all. If unread messages remain, filtered reads report the count and
an unfiltered-read hint on stderr, including when no messages match.
Address one or several members:
runner msg post --to coder --to reviewer "please check"
runner msg post "@coder @reviewer please check"
Explicit --to wins; mid-text mentions do not address. Unknown handles and @human are refused.
Reply to the sender. Broadcast only when everyone must act. Do not post receipt-only replies
or standby messages; consuming your inbox with runner msg read is still required.
START AND WATCH (REQUIRED)
After every successful runner mission start, arm runner mission feed <id> --follow --json
for the exact returned ID in your host's supported background/watch facility before
reporting delegation complete. Keep exactly one watcher per mission; reuse it on later
turns. Events, stderr diagnostics and process exit must reach you during later work and
while idle. An unread background log or a completion-only notification is not a watch.
Plain mission start still returns its result and exits; it does not block on watching.
Surface handoffs, human questions, completion, stop/crash and lost connections promptly.
The daemon pushes new events; each event is flushed immediately, without notification batching.
Routine busy/idle and inbox noise stays hidden by default; do not add --all for a watch.
The follower ends on archive, completed/aborted mission state, or all sessions exiting.
Busy/idle and a crew message saying "done" do not end a live mission. Resume needs a new watch.
A watch request timeout after 30 seconds exits 1; Runner may still be running.
Each JSON event line carries next_offset, the cursor that resumes after that event.
Clean up on exit. On host expiry or unexpected exit, report the gap, check mission show,
and re-arm one watch if still active with --since <next_offset of the last event line
you received> --oldest-first; a watch error prints its own resume command. Only when no event
line arrived, re-arm with --since 0 --oldest-first and skip event IDs already seen.
Never silently abandon a watch. Stop any undeliverable follower before ending your turn.
If your host cannot deliver background events, tell the user plainly:
"Mission <id> started, but this host cannot notify me of later events;
automatic watching is unavailable. Follow with: runner mission feed <id> --follow --json"
Do not claim monitoring is active without a verified delivery path.
WATCH FACILITIES
A facility qualifies only if its own documentation says each output line and the exit
reach you while idle. Claude Code's Monitor does: run the feed command with 2>&1 so stderr
diagnostics notify you too, set timeout_ms to its maximum, and re-arm on expiry as above.
A completion-only background command, a PTY session you must poll, a task list or a log
file does not qualify. Without a qualifying facility, take the limitation path above.
Do not install extensions or change agent configuration to create one.
IDENTITY
Inside a mission, mission commands carry the caller's own handle.
Outside a mission, --as <handle> names a roster slot and is refused if that handle is not in the roster.
Outside a mission with no --as, the caller is the person at the app.
Do not use --as to speak as another agent's slot; it exists for a slot you hold.
runner
Operate Runner missions, crews, roles, projects, and chats.
Usage: runner [OPTIONS] <COMMAND>
Commands:
daemon Inspect or stop the local daemon
status Check whether Runner is running
project Manage projects
role Manage reusable agent roles
crew Manage crews and their slots
mission Manage crew missions
chat Start direct chats
session Inspect and manage sessions
msg Post to or read a mission log
signal Emit a typed signal
ask Ask the person with --human
call Call any registered socket tool
help Print the command reference
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
-V, --version Print version
runner daemon
Inspect or stop the local daemon
Usage: runner daemon [OPTIONS] <COMMAND>
Commands:
status
stop
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner daemon status
Usage: runner daemon status [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner daemon stop
Usage: runner daemon stop [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner status
Check whether Runner is running
Usage: runner status [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner project
Manage projects
Usage: runner project [OPTIONS] <COMMAND>
Commands:
list List projects
show Show one project by ID or exact name
create Create a project
rename Rename a project
delete Delete a project and archive its members
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner project list
List projects
Usage: runner project list [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner project show
Show one project by ID or exact name
Usage: runner project show [OPTIONS] <PROJECT>
Arguments:
<PROJECT>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner project create
Create a project
Usage: runner project create [OPTIONS] <NAME>
Arguments:
<NAME>
Options:
--json Print the tool result as exact JSON
--path <PATH> Project directory; defaults to the current directory
-q Print only result IDs
-h, --help Print help
runner project rename
Rename a project
Usage: runner project rename [OPTIONS] <PROJECT> <NAME>
Arguments:
<PROJECT>
<NAME>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner project delete
Delete a project and archive its members
Usage: runner project delete [OPTIONS] <PROJECT>
Arguments:
<PROJECT>
Options:
--force Allow deletion when the project has running members
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner role
Manage reusable agent roles
Usage: runner role [OPTIONS] <COMMAND>
Commands:
list List reusable roles
show Show a role by handle
create Create a role
update Update a role
delete Delete a role
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner role list
List reusable roles
Usage: runner role list [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner role show
Show a role by handle
Usage: runner role show [OPTIONS] <HANDLE>
Arguments:
<HANDLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner role create
Create a role
Usage: runner role create [OPTIONS] <HANDLE>
Arguments:
<HANDLE>
Options:
--json Print the tool result as exact JSON
--name <NAME> Display name
-q Print only result IDs
--runtime <RUNTIME> Agent runtime registry name
--model <MODEL> Model override; pass an empty value to clear it
--effort <EFFORT> Reasoning effort; pass an empty value to clear it
--speed <SPEED> Codex Speed: inherit, standard, or fast [possible values: inherit, standard, fast]
--permission <PERMISSION> Permission mode: default, accept_edits, auto, or bypass
--prompt <PROMPT> Inline system prompt; pass an empty value to clear it
--prompt-file <PROMPT_FILE> Read the system prompt from a file, or - for stdin
--arg <ARGS> Append one runtime argument; repeat for multiple arguments
--env <ENV> Set one KEY=VALUE environment entry; repeat as needed
--cwd <CWD> Working directory; pass an empty value to clear it
-h, --help Print help
runner role update
Update a role
Usage: runner role update [OPTIONS] <HANDLE>
Arguments:
<HANDLE>
Options:
--json Print the tool result as exact JSON
--name <NAME> Display name
-q Print only result IDs
--runtime <RUNTIME> Agent runtime registry name
--model <MODEL> Model override; pass an empty value to clear it
--effort <EFFORT> Reasoning effort; pass an empty value to clear it
--speed <SPEED> Codex Speed: inherit, standard, or fast [possible values: inherit, standard, fast]
--permission <PERMISSION> Permission mode: default, accept_edits, auto, or bypass
--prompt <PROMPT> Inline system prompt; pass an empty value to clear it
--prompt-file <PROMPT_FILE> Read the system prompt from a file, or - for stdin
--arg <ARGS> Append one runtime argument; repeat for multiple arguments
--env <ENV> Set one KEY=VALUE environment entry; repeat as needed
--cwd <CWD> Working directory; pass an empty value to clear it
-h, --help Print help
runner role delete
Delete a role
Usage: runner role delete [OPTIONS] <HANDLE>
Arguments:
<HANDLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew
Manage crews and their slots
Usage: runner crew [OPTIONS] <COMMAND>
Commands:
list List crews
show Show one crew and its slots
create Create a crew
update Update a crew
delete Delete a crew
add Add a role to a crew
set Update one crew slot by handle
remove Remove a slot from a crew
lead Make a slot the crew lead
order Reorder every crew slot by handle
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew list
List crews
Usage: runner crew list [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew show
Show one crew and its slots
Usage: runner crew show [OPTIONS] <CREW>
Arguments:
<CREW>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew create
Create a crew
Usage: runner crew create [OPTIONS] <NAME>
Arguments:
<NAME>
Options:
--conventions-file <CONVENTIONS_FILE> Read crew conventions from a file, or - for stdin
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew update
Update a crew
Usage: runner crew update [OPTIONS] <CREW>
Arguments:
<CREW>
Options:
--json Print the tool result as exact JSON
--name <NAME> New crew name
--conventions-file <CONVENTIONS_FILE> Read crew conventions from a file, or - for stdin
-q Print only result IDs
-h, --help Print help
runner crew delete
Delete a crew
Usage: runner crew delete [OPTIONS] <CREW>
Arguments:
<CREW>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew add
Add a role to a crew
Usage: runner crew add [OPTIONS] <CREW> <ROLE>
Arguments:
<CREW>
<ROLE>
Options:
--as <HANDLE> Assign this in-crew handle
--json Print the tool result as exact JSON
-q Print only result IDs
--runtime <RUNTIME> Override the role runtime for this slot
--model <MODEL> Override the role model for this slot
--effort <EFFORT> Override the role effort for this slot
--speed <SPEED> Codex Speed override: inherit, standard, or fast [possible values: inherit, standard, fast]
-h, --help Print help
runner crew set
Update one crew slot by handle
Usage: runner crew set [OPTIONS] <CREW> <HANDLE>
Arguments:
<CREW>
<HANDLE>
Options:
--as <NEW_HANDLE> Rename the in-crew handle
--json Print the tool result as exact JSON
-q Print only result IDs
--runtime <RUNTIME> Override the runtime; pass an empty value to inherit
--model <MODEL> Override the model; pass an empty value to inherit
--effort <EFFORT> Override the effort; pass an empty value to inherit
--speed <SPEED> Codex Speed override: inherit, standard, or fast [possible values: inherit, standard, fast]
-h, --help Print help
runner crew remove
Remove a slot from a crew
Usage: runner crew remove [OPTIONS] <CREW> <HANDLE>
Arguments:
<CREW>
<HANDLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew lead
Make a slot the crew lead
Usage: runner crew lead [OPTIONS] <CREW> <HANDLE>
Arguments:
<CREW>
<HANDLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner crew order
Reorder every crew slot by handle
Usage: runner crew order [OPTIONS] <CREW> <HANDLES>...
Arguments:
<CREW>
<HANDLES>...
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission
Manage crew missions
Usage: runner mission [OPTIONS] <COMMAND>
Commands:
list List active missions
show Show an operational mission snapshot
start Start a mission
add-role Add a configured worker role to a running mission (person only)
remove-role Remove a worker from a running mission (person only)
stop Stop a running mission
resume Resume every stopped session in a mission
archive Archive a mission
unarchive Restore an archived mission to active lists
rename Rename a mission
pin Pin a mission
unpin Unpin a mission
move Move a mission into a project or unfile it
feed Print a window from the mission event feed
answer Answer a pending person question
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission list
List active missions
Usage: runner mission list [OPTIONS]
Options:
--crew <CREW> Filter by crew ID or exact name
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission show
Show an operational mission snapshot
Usage: runner mission show [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission start
Start a mission
Usage: runner mission start [OPTIONS]
Options:
--crew <CREW> Crew ID or exact name
--json Print the tool result as exact JSON
--goal <GOAL> Mission goal text
-q Print only result IDs
--goal-file <GOAL_FILE> Read the mission goal from a file, or - for stdin
--title <TITLE> Mission title; defaults from the goal or crew name
--project <PROJECT> Project ID or exact name whose directory the mission uses
--cwd <CWD> Mission directory; defaults to the current directory
-h, --help Print help
runner mission add-role
Add a configured worker role to a running mission (person only)
Usage: runner mission add-role [OPTIONS] <MISSION> <ROLE>
Arguments:
<MISSION>
<ROLE>
Options:
--handle <HANDLE>
--json Print the tool result as exact JSON
-q Print only result IDs
--runtime <RUNTIME>
--model <MODEL>
--effort <EFFORT>
--speed <SPEED> [possible values: inherit, standard, fast]
--task <TASK>
--task-file <TASK_FILE>
-h, --help Print help
runner mission remove-role
Remove a worker from a running mission (person only)
Usage: runner mission remove-role [OPTIONS] <MISSION> <HANDLE>
Arguments:
<MISSION>
<HANDLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission stop
Stop a running mission
Usage: runner mission stop [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission resume
Resume every stopped session in a mission
Usage: runner mission resume [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission archive
Archive a mission
Usage: runner mission archive [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission unarchive
Restore an archived mission to active lists
Usage: runner mission unarchive [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission rename
Rename a mission
Usage: runner mission rename [OPTIONS] <MISSION> <TITLE>
Arguments:
<MISSION>
<TITLE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission pin
Pin a mission
Usage: runner mission pin [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission unpin
Unpin a mission
Usage: runner mission unpin [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner mission move
Move a mission into a project or unfile it
Usage: runner mission move [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--json Print the tool result as exact JSON
--project <PROJECT> Destination project ID or exact name
-q Print only result IDs
--unfile Remove the mission from its project
-h, --help Print help
runner mission feed
Print a window from the mission event feed
Usage: runner mission feed [OPTIONS] [MISSION]
Arguments:
[MISSION]
Options:
--follow Follow pushed events until the mission ends or all its sessions exit
--json Print the tool result as exact JSON
-q Print only result IDs
--since <SINCE> Start after this byte offset
--limit <LIMIT> Maximum events to return
--oldest-first Ask the backend for chronological order
--types <TYPES> Comma-separated signal kinds (or message) to include
--from <FROM> Include only events from this handle
--all Include session_status and inbox_read noise
-h, --help Print help
runner mission answer
Answer a pending person question
Usage: runner mission answer [OPTIONS] <MISSION> <QUESTION_ID> <CHOICE>
Arguments:
<MISSION>
<QUESTION_ID>
<CHOICE>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner chat
Start direct chats
Usage: runner chat [OPTIONS] <COMMAND>
Commands:
start Start a role-backed or runtime-only direct chat
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner chat start
Start a role-backed or runtime-only direct chat
Usage: runner chat start [OPTIONS] [ROLE]
Arguments:
[ROLE]
Options:
--json Print the tool result as exact JSON
--runtime <RUNTIME> Start a role-free chat with this runtime
--model <MODEL> Model override
-q Print only result IDs
--effort <EFFORT> Reasoning effort override
--speed <SPEED> Codex Speed override: inherit, standard, or fast [possible values: inherit, standard, fast]
--project <PROJECT> Project ID or exact name whose directory the chat uses
--cwd <CWD> Chat directory; defaults to the current directory
-h, --help Print help
runner session
Inspect and manage sessions
Usage: runner session [OPTIONS] <COMMAND>
Commands:
list List recent direct-chat sessions
show Show one session with its live agent status
stop Stop a session and leave it resumable
archive Stop and archive a direct chat
resume Resume a stopped session
restart Restart a mission session with a fresh conversation
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session list
List recent direct-chat sessions
Usage: runner session list [OPTIONS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session show
Show one session with its live agent status
Usage: runner session show [OPTIONS] <SESSION>
Arguments:
<SESSION>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session stop
Stop a session and leave it resumable
Usage: runner session stop [OPTIONS] <SESSION>
Arguments:
<SESSION>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session archive
Stop and archive a direct chat
Usage: runner session archive [OPTIONS] <SESSION>
Arguments:
<SESSION>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session resume
Resume a stopped session
Usage: runner session resume [OPTIONS] <SESSION>
Arguments:
<SESSION>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner session restart
Restart a mission session with a fresh conversation
Usage: runner session restart [OPTIONS] <SESSION>
Arguments:
<SESSION>
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner msg
Post to or read a mission log
Usage: runner msg [OPTIONS] <COMMAND>
Commands:
post Post a mission message
read Read new inbox messages; --all shows crew history since joining
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner msg post
Post a mission message
Usage: runner msg post [OPTIONS] <TEXT>
Arguments:
<TEXT>
Options:
--json Print the tool result as exact JSON
--to <TO> Address a roster handle; repeat for several recipients
--mission <MISSION> Mission ID or unique prefix; required outside a mission
-q Print only result IDs
--as <FROM> Attribute the message to this roster handle
-h, --help Print help
runner msg read
Read new inbox messages; --all shows crew history since joining
Usage: runner msg read [OPTIONS]
Options:
--all Show all crew messages since joining
--json Print the tool result as exact JSON
-q Print only result IDs
--since <SINCE> Return messages after this event ID without marking them read
--from <FROM> Filter by sender handle without marking messages read
--mission <MISSION> Mission reference; only the caller's mission is supported
-h, --help Print help
runner signal
Emit a typed signal
Usage: runner signal [OPTIONS] <TYPE>
Arguments:
<TYPE> Signal type from the Runner coordination protocol
Options:
--json Print the tool result as exact JSON
--payload <PAYLOAD> JSON object carried by the signal
--mission <MISSION> Mission ID or unique prefix; required outside a mission
-q Print only result IDs
--as <FROM> Attribute the signal to this roster handle
-h, --help Print help
runner ask
Ask the person with --human
Usage: runner ask [OPTIONS] [QUESTION]
Arguments:
[QUESTION] Question for the mission lead
Options:
--context <CONTEXT> Legacy context flag; direct lead questions use msg post --to
--json Print the tool result as exact JSON
--human <HUMAN> Ask the person at the app
-q Print only result IDs
--choices <CHOICES> Comma-separated choices for --human
--mission <MISSION> Mission ID or unique prefix; required outside a mission
--as <FROM> Attribute the question to this roster handle
-h, --help Print help
runner call
Call any registered socket tool
Usage: runner call [OPTIONS] <TOOL> [ARGUMENTS]
Arguments:
<TOOL>
[ARGUMENTS]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help
runner help
Print the command reference
Usage: runner help [OPTIONS] [TOPIC]
Arguments:
[TOPIC]
Options:
--json Print the tool result as exact JSON
-q Print only result IDs
-h, --help Print help