Start and manage long-running background services such as development servers. Check their status and output logs, verify listening ports, retrieve local URLs, and restart services while keeping the same service ID.
Daemon process management — start, monitor, restart, stop long-running services. ALWAYS invoke this skill when the user asks to start, check, restart, or stop background services, dev servers, or daemon processes.
Review the instructions the agent will receive before you install.
yao_daemon_start — Start a long-running service. Always runs in background; returns daemon_id immediately.yao_daemon_list — List daemons. Supports --status filter and --limit/--offset pagination.yao_daemon_status — Get full details for a daemon (daemon_port, daemon_url, started_at, daemon_restarts).yao_daemon_stop — Stop a running daemon.yao_daemon_restart — Restart a daemon (preserves the same ID, increments daemon_restarts).yao_job_output — Read daemon output (pass daemon_id as job_id).| Param | Required | Default | Description |
|---|---|---|---|
command | yes | — | Shell command to execute |
description | yes | — | Concise 5–10 word summary |
port | no | auto-detect | Expected listen port; the system verifies the port is ready before reporting it |
args | no | — | Additional arguments |
working_dir | no | workspace | Working directory |
| Param | Default | Description |
|---|---|---|
status | all | Filter: running, stopping, completed, failed, cancelled, port_detecting, all |
limit | 20 | Max items per page |
offset | 0 | Skip first N items |
| Param | Required | Description |
|---|---|---|
daemon_id | yes | The daemon ID |
All responses are JSON. Daemon objects contain:
id, kind, session_id, status, description,
command, command_argv, command_line, pid, pgid, user,
started_at, finished_at, duration_ms, exit_code, signal, detail,
daemon_port, daemon_url, daemon_restarts, seq
daemon_port — The verified listen port (0 if not yet detected or not specified).daemon_url — http://localhost:<port> (empty until port is verified as listening by the daemon's own process tree).daemon_restarts — Number of restarts via yao_daemon_restart.started_at — Unix millisecond timestamp of the most recent start.Port detection behavior:
Both modes start by returning port_detecting with daemon_port: 0 if the port is not confirmed within the initial ~2s settle window. If the port is confirmed within ~2s, start returns running with daemon_port/daemon_url directly.
--port N (explicit): Verifies port N is listening AND owned by this daemon's process tree. If not confirmed within ~2s, returns status: "port_detecting". Poll yao_daemon_status until status changes to running — either with daemon_port: N and daemon_url (port confirmed), or with daemon_port: 0 and detail: "specified port N not detected within timeout" (~60s).--port (auto-detect): Scans the daemon's process tree for any listening TCP port. If not found within ~2s, returns status: "port_detecting". The scan continues in background (~60s); if a port is found, status changes to running with daemon_port/daemon_url and a <BackgroundJobReceipt> is delivered. If no port is found after ~60s, status changes to running with daemon_port: 0.{ "daemons": [...], "total": 5, "offset": 0, "limit": 20, "has_more": false }
{ "job": { ... }, "detail": "already in terminal state (cancelled); no action taken" }
Structured JSON with code and message:
{ "error": { "code": "not_found", "message": "daemon \"dmn_xxx\" not found" } }
Codes: invalid_argument, not_found, internal.
description.port when starting a server — the system confirms the port is actually listening before reporting daemon_url.yao_daemon_stop to terminate.yao_daemon_restart to restart a crashed or misconfigured daemon without losing its ID.yao_job_output passing the daemon_id as job_id.When a daemon's port becomes ready, the system delivers a <BackgroundJobReceipt> XML message containing the daemon_url. Use that URL for subsequent HTTP requests to the daemon.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Manage agents on a host by listing their capabilities, reading source code, and downloading or deploying agents in the smith namespace. Check available language model connections and their roles without exposing API keys.
Turn recorded speech into text from supported audio files such as mp3, m4a, and wav. Detect the language automatically or specify it, and list available speech recognition providers and models before choosing one.
Check kanban boards, which organize tasks into columns, and retrieve task lists filtered by status, assistant, or board. Use the returned data to see task counts and review work in progress.
Classify messages, rate content, or estimate probabilities using defined question types. Receive machine-readable answers, with confidence scores for classification and ratings, and apply the same questions to multiple inputs in a batch.
Look up callable processes in Yao, inspect their arguments and return types, and check whether a process name is valid. Use the documentation to prepare calls and find suggested alternatives for misspelled names.
Describe photos, screenshots, and charts in text, generate images from written prompts, or edit existing images. Supports background replacement, transparent output, and edits limited to an area defined by a mask.