Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
CRONSTABLE_WEB_TOKENNoBearer token the `cronstable mcp` stdio bridge sends to the cronstable HTTP API. Read from the CRONSTABLE_WEB_TOKEN environment variable. Optional when the daemon's API is unauthenticated (the default), but required when the daemon sets `web.authToken`.

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

cronstable's MCP server. Read-only 'observe' tools describe jobs, workflows, cluster health, metrics and saved state. Mutating tools (run/cancel/pause/resume a job, run/backfill/approve a workflow) require confirm=true and appear only when the operator disabled readOnly and the presented token allows them. Start with cron_get_status or cron_list_jobs. When authoring a schedule, verify it with cron_validate_schedule / cron_explain_schedule (the server's scheduling engine) before proposing it; cron_why_no_run explains why a job's schedule did or did not match a given timestamp.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "listChanged": false
}
completions
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
cron_get_statusA

Show each job's status: running, disabled, or scheduled, one page at a time. Use it for a quick check of every job; use cron_list_jobs for schedules, next runs, and last outcomes, or cron_get_job for one job's details.

cron_list_jobsA

List jobs with schedule, enabled/running state, next run, and last outcome, one page at a time. Use it to find a job's exact name or to survey schedules. Each row has the same fields as cron_get_job, which fetches one job by name.

cron_get_jobA

Show details for one job (schedule, command, last run, live resources, retry/slot state). Use it when you know the job's name; use cron_list_runs for its run history, or cron_tail_job_logs for its output.

cron_list_runsA

Show saved run history, success rate, and duration statistics for one job, newest run last. Use it to inspect individual runs; use cron_get_job_trends for success rates over fixed windows, or cron_why_no_run when an expected run is missing.

cron_get_job_trendsA

Per-window (1h/24h/7d/30d/all) success-rate and duration statistics from the saved run history for one job. Use it to see whether a job fails more often or runs longer than it used to; use cron_list_runs for the runs behind the numbers.

cron_get_job_resourcesA

CPU/RSS time series for a job's live and recent runs (monitorResources jobs), plus its resourceLimits caps and each capped run's limits record, which says whether the caps applied and whether the run hit its memory cap. Use it for one job's resource use; use cron_get_node for the whole host, or cron_query_metrics for Prometheus samples.

cron_get_clusterA

This node's cluster/leadership view (peers, quorum, role, live load). Returns enabled:false if no cluster is configured. Use it to see which node leads and whether quorum holds; use cron_get_fleet to compare each job's runs across nodes.

cron_get_fleetA

Run status for each job on each node. Returns enabled:false when cluster peer communication is unavailable. Use it to check where each job last ran and how it ended; use cron_get_cluster for leadership and quorum.

cron_get_nodeA

This node's live whole-host CPU/memory (optionally with the recent history). Use it to check host load; use cron_get_job_resources for one job's share.

cron_query_metricsA

Parsed samples from the Prometheus /metrics exposition, optionally filtered by metric name. Use it for counters across all jobs, such as run outcomes, retries, and SLA breaches; use cron_get_job_trends for one job's success rate.

cron_get_versionA

Show the server version, job-set ID, and job count. Use it to identify the daemon, or to compare job-set IDs across replicas, which match when the replicas' jobs agree on the configuration fields the ID covers, even if secret values or host-specific settings differ; use cron_get_cluster for the replicas' roles.

cron_tail_job_logsA

Last retained stdout/stderr lines of a job's newest running instance, or of its last finished run when none is running, with a cursor to poll for newly appended lines (the poll form of the live log stream). Use it to read what a job printed; use cron_list_runs for exit status and timing.

cron_schedule_pressureA

Upcoming runs for enabled schedules over the next hours (default 24, max 168), grouped by hour and minute in tz (default UTC). Shows when jobs are scheduled together and which minute positions have no scheduled runs. Use it to see how busy each minute is; use cron_suggest_slot to pick a time for a new job.

cron_schedule_duplicatesA

Jobs with identical run times (for example, */5 and 0-59/5 in the same time zone), among enabled jobs and workflow schedules. Use this to find schedules that could be spread out; use cron_schedule_pressure for different schedules that share some minutes.

cron_suggest_slotA

Suggest a time with fewer scheduled runs in the next 24 hours, as a cron expression with two alternatives and the busiest time. Use it when adding a job, then check the result with cron_validate_schedule.

cron_validate_scheduleA

Check a cron expression before saving it as a job: valid true/false with the engine's exact error (including wrong-field hints for Quartz-style forms), the plain-English description, the normalized form, advisory warnings, and the first scheduled run, all from the server's scheduling engine. The dialect includes L (last day), L-n (n days before it), nW / LW (nearest / last weekday) and Ln / d#n (last / nth weekday: L5 = last Friday, 5#3 = third Friday). Use it before saving a new or changed schedule; use cron_explain_schedule to list more upcoming runs, or cron_why_no_run for a job that already exists.

cron_explain_scheduleA

Decode a cron expression into a plain-English description, its next count run times as ISO timestamps in tz, and advisory warnings from the server's scheduling engine. Use it to read an expression or preview its runs; use cron_validate_schedule for the check before saving, or cron_why_no_run to test an existing job at a timestamp.

cron_why_no_runA

Explain field-by-field why a job's schedule did or did not select a timestamp ('day-of-week Tuesday is not in Monday and Friday'), with the previous and next scheduled runs and notes on this dialect's day-field AND rule and DST effects. Use it when an expected run is missing. If the schedule does match, the answer refers to execution history (cron_list_runs) instead.

cron_list_poolsA

Show each pool's capacity, queued work, and recent queue outcomes. Use it when jobs wait for pool capacity, or to find the queue entry ID that cron_cancel_queued takes when the act tools are on.

Prompts

Interactive templates invoked by user choice

NameDescription
triage_job_failureFind why a job failed using its runs, trends, logs, and host health.
blast_radiusIdentify jobs and workflows affected by a failing job or workflow.
fleet_health_summaryA summary of cluster, node, and job health.

Resources

Contextual data attached and managed by the client

NameDescription
statusLive status of every job.
clusterThis node's cluster/leadership view.
fleetThe cluster-wide jobs x nodes matrix.
versionShow the server version, job-set ID, and job count.

TDQS

A4.1/5.0

Scored across 19 tools

Disambiguation4/5

Nearly every description explicitly names its sibling tools and states when to prefer them (e.g. cron_get_job_resources vs cron_get_node vs cron_query_metrics), which makes the set unusually easy to navigate. The remaining overlap is real but small: cron_validate_schedule and cron_explain_schedule both decode an expression and list upcoming runs, and cron_get_status vs cron_list_jobs differ mainly in emphasis.

Naming Consistency4/5

All 19 tools share the cron_ prefix and mostly follow a verb_noun shape (cron_get_job, cron_list_runs, cron_validate_schedule, cron_suggest_slot). A few break the pattern with noun or question forms (cron_schedule_pressure, cron_schedule_duplicates, cron_why_no_run), which are readable but less predictable.

Tool Count4/5

19 tools is on the heavy side, but the domain genuinely spans three areas — job/run inspection, schedule authoring helpers, and cluster/host observability — and each tool covers a distinct slice of those. No tool looks like filler, though the observability tail (fleet, node, cluster, metrics, version) is where trimming would be easiest.

Completeness3/5

The read/diagnostic surface is thorough (status, details, runs, trends, resources, logs, schedule math, cluster view), but the entire mutation lifecycle is absent: no create/update/delete job, no enable/disable, no run-now. cron_list_pools even points at a cron_cancel_queued tool that is not in this set, so agents will hit a dead end when they try to act rather than observe.

Maintenance

ActivityNo data
ResponsivenessOutdated data