# avr job Inspect Avrea job VMs (SSH, metrics, logs). ```sh avr job [OPTIONS] COMMAND [ARGS]... ``` ## Subcommands ### `avr job list` List jobs for an organization. ```sh avr job list [OPTIONS] ``` ```sh Examples: avr job list avr job list --status failure --limit 5 avr job list --status in_progress --json job_name,state,conclusion avr job list --since 24h avr job list --json '?' # list available fields avr job list --json '*' # all fields ``` ```sh JSON FIELDS completed_at, conclusion, created_at, duration_seconds, platform_job_id, platform_run_id, job_id, job_name, labels, repository, repository_id, running_on_avrea, started_at, state ``` **Options** - --org <TEXT> — Organization ID. Uses default org if not specified (see: avr config set org). - --repo <TEXT> — Filter by repository (org/repo or rep-xxx, repeatable). Auto-detected from git remote if omitted. _(repeatable)_ - --name <TEXT> — Filter by job name (repeatable). _(repeatable)_ - --status <CHOICE> — Filter by state (queued, in_progress, completed) or conclusion (success, failure, ...). Repeatable. _(choices: `action_required`, `cancelled`, `completed`, `failure`, `in_progress`, `neutral`, `queued`, `skipped`, `stale`, `startup_failure`, `success`, `timed_out` · repeatable)_ - --on-avrea / --shadowing — Filter by Avrea-run vs shadowing jobs. - -w, --workflow <TEXT> — Filter by workflow ID (wfl-xxx, repeatable). _(repeatable)_ - --since <TEXT> — Relative time window: '7d', '24h', etc. - -L, --limit <INTEGER RANGE> — Max jobs to return. _(default: `20`)_ - --cursor <TEXT> — Pagination cursor from a previous response. - --order <CHOICE> — Sort order. _(choices: `created_at.desc`, `created_at.asc` · default: `created_at.desc`)_ - --json <TEXT> — Output JSON. Pass comma-separated field names, "*" for all fields, or "?" to list available fields. - -q, --jq <TEXT> — Filter --json output through a jq expression. ### `avr job logs` View logs for a job, grouped by step. ```sh avr job logs [OPTIONS] JOB_ID ``` ```sh Examples: avr job logs job-abc123 avr job logs job-abc123 --failed avr job logs job-abc123 --step "Build" --level error avr job logs job-abc123 --follow ``` **Arguments** - JOB_ID **Options** - --org <TEXT> — Organization ID. - --failed — Only show logs from failed steps. - --step <TEXT> — Filter to a specific step by name. - --level <CHOICE> — Filter by log level. _(choices: `debug`, `info`, `notice`, `warning`, `error`)_ - --follow, -f — Follow logs for in-progress jobs. - --all-levels — Include diagnostic-level lines (hidden by default). - --no-pager — Print directly to stdout instead of paging through `less`. Same as setting AVR_PAGER=''. ### `avr job metrics` Show CPU/memory/IO gauges for a job's VM. ```sh avr job metrics [OPTIONS] JOB_ID ``` ```sh Examples: avr job metrics job-abc123 avr job metrics job-abc123 --source cpu --source network avr job metrics job-abc123 --watch ``` **Arguments** - JOB_ID **Options** - --org <TEXT> — Organization ID. - --source <CHOICE> — Metric source (repeatable). Defaults to cpu and memory. _(choices: `cpu`, `memory`, `filesystem`, `load`, `disk-io`, `disk-ops`, `network` · repeatable)_ - --start <INTEGER> — Start time (Unix seconds). Defaults to execution start. - --end <INTEGER> — End time (Unix seconds). Defaults to execution end or now. - -w, --watch — Refresh every 5 seconds (Ctrl-C to exit). - --json — Output raw metrics responses as JSON. ### `avr job ssh` SSH into a running job's VM. ```sh avr job ssh [OPTIONS] JOB_ID ``` **Arguments** - JOB_ID **Options** - --print-command — Print the SSH command instead of connecting. - --show-password — Display the SSH password (use with caution). ### `avr job view` View a single job with its steps. ```sh avr job view [OPTIONS] JOB_ID ``` ```sh Examples: avr job view job-abc123 avr job view job-abc123 --log-failed avr job view job-abc123 --json conclusion,steps avr job view job-abc123 --json '*' --jq '.steps[] | select(.conclusion=="failure")' ``` ```sh JSON FIELDS completed_at, conclusion, created_at, duration_seconds, platform_job_id, platform_run_id, job_id, job_name, labels, repository, repository_id, running_on_avrea, started_at, state, steps, workflow_run ``` **Arguments** - JOB_ID **Options** - --org <TEXT> — Organization ID. - --log — Print full logs for the job. - --log-failed — Print logs only for failed steps. - --json <TEXT> — Output JSON. Pass comma-separated field names, "*" for all fields, or "?" to list available fields. - -q, --jq <TEXT> — Filter --json output through a jq expression. - --web — Open in browser. - --no-pager — Print logs directly to stdout instead of paging through `less`. Same as setting AVR_PAGER=''. ### `avr job watch` Watch active jobs with auto-refresh (Ctrl+C to stop). ```sh avr job watch [OPTIONS] ``` **Options** - --org <TEXT> — Organization ID. Uses default org if not specified. - --repo <TEXT> — Filter by repository (org/repo or rep-xxx, repeatable). Auto-detected from git remote if omitted. _(repeatable)_ - --name <TEXT> — Filter by job name (repeatable). _(repeatable)_ - --interval <INTEGER> — Refresh interval in seconds. _(default: `5`)_ - --ndjson — Emit one JSON object per refresh (default when stdout isn't a TTY).