List all tasks
const url = 'http://localhost:9477/api/tasks';const options = {method: 'GET'};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request GET \ --url http://localhost:9477/api/tasksResponses
Section titled “Responses”OK
object
For services: whether instances start at boot. False boots in the stopped state until started via API/UI.
What to do when cron ticks are missed during downtime
Provenance metadata for tasks imported from a docker compose file
object
For services: service names that must be healthy before this one starts at boot — boot ordering only, not a workflow DAG
Environment variables overlaid on the task’s process env. Values are visible to authenticated operators in the API/UI; env_file values merge in beneath the inline entries.
object
What the run’s environment starts from: ‘inherit’ (the daemon’s, the default) or ‘clean’ (PATH, SHELL, HOME, USER/LOGNAME only, as crond gives a job)
Path to a dotenv file whose KEY=VALUE pairs merge into env (inline entries win). Values are visible in the API/UI like inline env.
Process exit codes treated as success; defaults to [0]
Window between the stop signal and SIGKILL when a run is stopped, in nanoseconds
For services: an instance that runs at least this long counts as healthy — resets the restart counter and clears the failed-start streak; fast exits below it count toward restart_attempts, in nanoseconds
Why this task is loaded but not on the scheduler: ‘cron’ means a live system cron daemon still reads the crontab it came from and is running it, so RunWisp stands down. Manual triggers still work. Empty means RunWisp owns the schedule.
For services: number of always-running instances
Cap how far a cron task’s start may slip so tasks sharing a fire time take turns through a daemon-wide one-at-a-time gate instead of stampeding; a run starts as soon as the gate frees and slips up to this window only under contention, in nanoseconds
Retention window in nanoseconds; 0 means no cap was configured
Row-count retention cap; 0 keeps no completed runs, omitted inherits the [defaults] value (or no cap)
Whether this is a scheduled task or an always-on service
Per-run log size cap in bytes
What to do when log output exceeds log_max_size
Cap on catch-up runs triggered when catch_up = all
Maximum overlapping runs allowed for this task
How overlapping runs are handled
Per-execution parameters an operator may supply at manual trigger time; scheduled runs use the declared defaults
object
When choices is set, allow values outside the list
Allowed values; renders as a dropdown
Default value used by scheduled runs and pre-filled in manual forms
Help text shown under the field
Canonical parameter key (env name, positional label, or option/flag token)
How the parameter renders into the run
Whether a manual trigger must supply a value
Value type; defaults to string
For services: boot start order, lowest first (name breaks ties). Start order only — not a dependency.
Maximum runs that can wait when on_overlap = queue
Whether and when a task is restarted after completion
For services: consecutive fast failures tolerated before an instance is marked FATAL and stops restarting
Backoff curve between consecutive restarts
Base delay before each restart, in nanoseconds
Backoff curve between consecutive retries
Base delay before each retry, in nanoseconds
For tasks: fire once at daemon startup, in addition to any cron schedule
Path to a dotenv file whose KEY=VALUE pairs are injected into the task’s process env. The path is visible in the API/UI; keys and values never leave the daemon.
Absolute path to the shell interpreter for run scripts; defaults to /bin/sh
Where this task’s definition came from: native (hand-authored TOML), staged (imported, not yet promoted), or cron (read live from a crontab via daemon.include_cron)
Absolute path of the crontab or staging file this task’s definition was read from; empty for hand-authored TOML
Signal sent to stop a run before SIGKILL; defaults to SIGTERM
Per-run timeout in nanoseconds
IANA timezone for cron evaluation; falls back to scheduler.timezone, then the daemon’s resolved system timezone
Octal file-creation mask applied to the run’s process; empty inherits the daemon’s umask
Run the process as this OS user, in ‘user’ or ‘user:group’ form (name or numeric id). Empty runs as the daemon’s user; switching users needs the daemon running as root.
Resolved working directory for the task’s process; empty inherits the daemon’s working directory. A literal “~” means the run-as user’s home, resolved at run time
Example
[ { "catchUp": "latest", "heldBy": "cron", "kind": "task", "logOnFull": "drop_new", "onOverlap": "queue", "parameters": [ { "kind": "env", "type": "string" } ], "restart": "never", "restartBackoff": "constant", "retryBackoff": "constant", "source": "staged", "stopSignal": "SIGTERM" }]default
Section titled “default”Error
object
A URL to the JSON Schema for this object.
A human-readable explanation specific to this occurrence of the problem.
Optional list of individual error details
object
Where the error occurred, e.g. ‘body.items[3].tags’ or ‘path.thing-id’
Error message text
The value at the given location
A URI reference that identifies the specific occurrence of the problem.
HTTP status code
A short, human-readable summary of the problem type. This value should not change between occurrences of the error.
A URI reference to human-readable documentation for the error.
Example
{ "$schema": "http://localhost:9477/schemas/ErrorModel.json", "detail": "Property foo is required but is missing.", "instance": "https://example.com/error-log/abc123", "status": 400, "title": "Bad Request", "type": "about:blank"}