Skip to main content
Glama

job_enqueue

Enqueue a validated job into the queue; intake reports every missing required field in one error, and host_choice='ask' leaves it unassigned.

Instructions

Enqueue a job (validated intake -> queued). The intake is the gate: every missing required field comes back in one error with a field list. Set host_choice='ask' to queue unassigned.

Input Schema

NameRequiredDescriptionDefault
kindYestest | demo | final
titleYesHuman name (the HUD strip / banner label)
branchNoBranch under test
commitNoCommit sha under test
originNoWho asked for this run: human | agent | driver | eval | probe | warmup (default agent). The learning loop reads the ledger by origin, so tag synthetic runs (driver/eval/probe/warmup) — an untagged repeat inflates a training artifact.
reasonNoWhy the run may exceed the 5-minute default
targetYesWhat is under test: feature/test name + branch@sha
cleanupNoThings that must be gone at the end
host_idNoHost to queue on, or set host_choice='ask'
purposeYesOne sentence: why this run exists
expectedYesPass criterion (an observable outcome)
watchingNoThe user is watching this run live
project_idNoProject this job belongs to
host_choiceNo'ask' = stays queued + unassigned until assigned
max_minutesNoCap; test/demo default 5, larger needs `reason`
screenshotsNoeach_step (default) | on_request | noneeach_step
requested_byNoCoordinator session id
needs_confirmNoTrue = the human iso_confirm gate decides pass/fail

Output Schema

NameRequiredDescriptionDefault
idYes
kindYes
linksNo
phaseNo
titleYes
eventsNo
originNoWho asked for this run (human | agent | driver | eval | probe | warmup) — scopes the learning loopagent
paramsNoIntake: purpose, expected, caps, flags, pending_shot, ...
statusYes
targetYes
workerNo
host_idNoNULL = not yet assigned (host_choice='ask')
summaryNoPresent once the job is closed out
web_urlNoAbsolute URL of this job's page in the web UI
positionNo
artifactsNo
created_atYes
project_idNo
started_atNo
updated_atYes
crash_countNocrash-level events (the queue board flags these)
error_countNoerror-level events (the queue board flags these)
finished_atNo
current_stepNoThe step awaiting a verdict (newest prompt event + whether it is answered)
requested_byNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A3.8/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the annotations it discloses two real behaviors: validation is aggregate (all missing required fields surface in a single error with a field list) and 'ask' leaves the job queued and unassigned until picked up. It does not spell out that repeated calls create duplicate jobs despite idempotentHint=false, which would have earned a 5.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two short sentences, front-loaded with the action and the state transition, then the two most operationally important facts. Nothing wasted.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given an output schema and 100% schema description coverage, the description needn't restate returns or parameter syntax, and it correctly focuses on the intake-gate behavior. For an 18-parameter tool it is a little thin on which fields the gate actually enforces, but it is sufficient to call correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3. The description's mention of host_choice='ask' and the 5-minute cap only restates what the schema already documents for those parameters; it adds no new semantics for the other 15 fields.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

A specific verb+resource ('enqueue a job') plus the state transition it performs ('validated intake -> queued'). An agent can distinguish it from the job_* siblings that consume/assign/finish jobs, though no sibling is named explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It hints at the creation context ('the intake is the gate') and gives one operational hint ('host_choice=ask to queue unassigned'), but there is no explicit when-to-use / when-not, no naming of alternatives like job_assign or job_next, and no prerequisite guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.