Ctrlk
For the complete documentation index, see llms.txt. This page is also available as Markdown.

jobs

Flow execution history and control — inspect past runs, monitor what's running now, download produced files, and pull errors for a specific run.

REST API: Jobs

OpenAPI spec: dist/job.yml

celigo jobs <subcommand> [args] [flags]

Supports all global flags.

Subcommands

Subcommand
Purpose

get <id>

Fetch one job; parent jobs auto-expand to the full family (parent + child jobs).

list

Last resort: list historical jobs with strict, index-backed filters.

current

List currently-running jobs.

cancel <id>

Cancel a running job.

diagnostics <id>

Fetch diagnostics for a job.

download-files <id>

List signed download URLs, or fetch a file's bytes (--file-id, -o).

purge-files <id>

Delete stored files for a job.

errors <id>

Open errors attributable to one job (walks the family to resolve scope).

resolved-errors <id>

Resolved errors attributable to one job (same resolution as errors).

run-stats

Flow run statistics.

celigo jobs get

Get a job by ID. Top-level run jobs (type=flow, retry, or bulk_retry) auto-expand to the full family — the run record with its child jobs inlined under children — because a run only carries aggregate counters and the per-step detail lives on the children. Child step jobs (type=export/import) return just the single record.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id.

Example

Corresponds to: GET /v1/jobs/{_id} — and, for parent jobs, GET /v1/jobs/{_id}/family.

celigo jobs list

Last resort: list historical jobs via the generic jobs endpoint with strict, index-backed filters. Prefer jobs current for running/active jobs, jobs get <id> for one run (children inlined), and jobs run-stats for aggregate history.

jobs list was restricted in celigo-cli 2026.8.4, at the platform team's request — free-form GET /v1/jobs queries were degrading the shared database for every account. --type is now required, only two filter shapes are supported, and the --export/--import flags were removed (child export/import jobs are not indexed by step ID — scope them by the parent run's job ID with --flow-job/--parent-job, or use jobs get <runId>, which inlines the child jobs). Every rejected combination fails client-side, before any network call, with the supported alternative named.

The two supported filter shapes:

--type

Required scope

flow or retry

--integration <id> or --flow <id>

export or import

--flow-job <id> or --parent-job <id> (the parent run's job ID)

Signature

Flags

Flag
Type
Default
Description

--type <type>

flow|retry|export|import

—

Required. Job type to list. (The long-defunct bulk_retry type is deliberately not accepted.)

--integration <id>

string

—

Scope --type flow/retry by _integrationId.

--flow <id>

string

—

Scope --type flow/retry by _flowId.

--flow-job <id>

string

—

Scope --type export/import by _flowJobId (the parent run's job ID).

--parent-job <id>

string

—

Scope --type export/import by _parentJobId.

--status <status>

completed|failed|queued|canceled|…

—

Filter by one status (the API has no multi-status filtering).

--created-gte <iso>

ISO 8601

—

createdAt >= lower bound. Additional narrowing only — not sufficient scope on its own.

--created-lte <iso>

ISO 8601

—

createdAt <= upper bound.

--started-gte <iso>

ISO 8601

—

startedAt >= lower bound.

--started-lte <iso>

ISO 8601

—

startedAt <= upper bound.

--limit <n>

integer

—

Stop after collecting this many rows (client-side early-stop).

Example

Corresponds to: GET /v1/jobs?... (filters are mapped to type, _integrationId, _flowId, _flowJobId, _parentJobId, status, createdAt_gte/lte, startedAt_gte/lte).

celigo jobs current

List currently-running jobs. Uses the POST /v1/jobs/current body-paginated endpoint.

Signature

Flags

Flag
Type
Default
Description

--integration <id>

string

—

Filter by integration ID (sent as _integrationIds).

--flow <id>

string

—

Filter by flow ID (sent as _flowIds).

--status <status>

comma-separated

—

One or more of queued, running, retrying, canceling.

--sandbox was removed in celigo-cli 2026.8.1. The environment comes from the profile's token, so a flag could contradict it. To query another environment, use a profile whose token targets it. See Profiles & regions.

Example

Corresponds to: POST /v1/jobs/current

celigo jobs cancel

Cancel a running job. Prompts for confirmation unless -y is passed.

⚠️ Canceling a running job cannot be undone. The job stops where it is; records already processed are kept, and the run does not resume.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id.

Flags

Flag
Type
Default
Description

-y, --yes

boolean

false

Skip the confirmation prompt.

Example

Corresponds to: PUT /v1/jobs/{_id}/cancel

celigo jobs diagnostics

Fetch diagnostics for a job (performance breakdown, retry metadata, etc.).

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id.

Example

Corresponds to: GET /v1/jobs/{_id}/diagnostics

celigo jobs download-files

List the signed download URLs for a job's files, or fetch a file's bytes to disk.

  • No flags — prints the signed download URLs for every file the job produced (it does not fetch the bytes).

  • --file-id — narrows to one file (e.g., an EDI transaction's s3Key).

  • -o — fetches the file's bytes and writes them to the given path.

Where do file ids come from? jobs get <id> returns a files[] array; each entry has an id (the storage file id), host (s3 or netsuite), and name. Pass that id to --file-id. For EDI specifically, the s3Key on an edi-transactions record is the file id.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id.

Flags

Flag
Type
Default
Description

--file-id <fileId>

string

—

Specific file id (from jobs get → files[].id, or an EDI s3Key).

-o, --output <path>

path

—

Fetch the file's bytes and write them to this path instead of stdout.

Example

Corresponds to: POST /v1/jobs/{_id}/files/signedURL — called with an empty body to list URLs, or { "fileIds": [...] } when --file-id is set. With -o, the CLI then GETs the returned signed URL to fetch the bytes.

celigo jobs purge-files

Delete stored files for a job. The job record itself is not removed. Prompts for confirmation unless -y is passed.

⚠️ Purging files is irreversible. The stored file artifacts are deleted permanently; the job record, diagnostics, and errors remain queryable.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id.

Flags

Flag
Type
Default
Description

-y, --yes

boolean

false

Skip the confirmation prompt.

Example

Corresponds to: DELETE /v1/jobs/{_id}/files

celigo jobs errors

Return open errors attributable to one job. The CLI resolves the job's flow and export/import steps from its family, then queries the flow-scoped error endpoint filtered by _flowJobId plus the job's run window. For resolved errors, use jobs resolved-errors.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id. Parent flow jobs fan out across their children; child jobs walk up to the parent to resolve _flowId.

Example

Corresponds to: GET /v1/flows/{_flowId}/{exportOrImportId}/errors filtered by _flowJobId and an occurredAt window — see the flow errors section of the REST API. Returns up to 1000 rows.

celigo jobs resolved-errors

Return resolved errors attributable to one job. Identical flow + step resolution as jobs errors, but queries the resolved-error endpoint with a resolvedAt window.

Signature

Arguments

Argument
Type
Required
Description

<id>

string

Yes

Job _id. Resolved the same way as jobs errors (parent fan-out / child walk-up).

Example

Corresponds to: GET /v1/flows/{_flowId}/{exportOrImportId}/resolved filtered by _flowJobId and a resolvedAt window. Returns up to 1000 rows.

celigo jobs run-stats

Flow run statistics, grouped by flow/integration.

Signature

Flags

Flag
Type
Default
Description

--integration <id>

string

—

Filter by integration ID (sent as _integrationIds).

--flow <id>

string

—

Filter by flow ID (sent as _flowIds).

--status <status>

comma-separated

—

One or more of completed, failed.

--sandbox was removed in celigo-cli 2026.8.1. Use a profile whose token targets the environment you want instead.

Example

Corresponds to: POST /v1/flows/runs/stats

Gotchas

  • jobs list is the last resort, and it enforces its own scope. --type is required, --type flow/retry must be scoped by integration or flow, and --type export/import by the parent run's job ID. A createdAt bound alone is not sufficient scope, and --status takes a single value. Prefer current, get, and run-stats — they answer most questions without the generic endpoint.

  • Child jobs have no _flowId. Export/import-type jobs carry _parentJobId (or _flowJobId) instead — which is also why jobs list scopes them by --flow-job/--parent-job rather than the removed --export/--import flags. jobs errors walks up to the parent automatically; if you query the flow errors API yourself, do the same walk first.

  • jobs errors requires a time window. The flow error endpoint rejects queries that lack occurredAt_gte/lte (or resolvedAt_gte/lte for jobs resolved-errors). The CLI derives this from startedAt/endedAt on the job — jobs that haven't ended yet fall back to "now" on the upper bound.

  • jobs current and jobs run-stats are POST body endpoints, not GET with a query string. A read-mode profile blocks them, because the client gates on the HTTP verb. Use an operate-mode profile for these two.

  • Environment comes from the profile, not a flag. --sandbox was removed in celigo-cli 2026.8.1. Every job command reads the environment that the profile's token targets. Add a second profile to query another environment.

  • jobs download-files without -o does not fetch bytes. It prints the signed URLs and returns — writing content to disk requires -o (optionally with --file-id to pick one file).

  • purge-files is not the same as cancelling or deleting a job. It removes the stored file artifacts only; the job record, diagnostics, and errors remain queryable.

Related

  • audit — who changed a resource (complement to jobs for runtime events).

  • flows — flows errors / flows resolve-errors operate on the step-level scope that jobs errors resolves into.

  • integrations — use jobs list --type flow --integration <id> to sweep runs across every flow in one integration.

Previousintegrations
Nextlookup-caches

Last updated

Was this helpful?