jev-triage Checked

For B, skip CLI/API calls. Mark agent_simulation or model_simulation, identify

the actual model when available, set jev_called: false, probability: null and

confidence: null. Return a value, evidence-based reason and needs_review; use

null/review when evidence is missing. Do not invent Jev output or probabilities.

Choice uses supplied labels, Noul uses booleans, Score uses integer rubric indices.

For A, use the existing jev-decide CLI with the chosen --provider openrouter

or --provider typesafe. If absent, explain the dependency; do not silently install.

--dry-run is offline validation, not a judgment. Exit 0 means selected/scored,

2 means review, 1 means error. Read each value: false Noul remains false. Selection

is not permission, and confidence is not accuracy. Keep unknown/review paths.

First request

Adapt the example (assets/example.json). The shared CLI needs Python 3.10+;

no sibling skill is needed. Host tools still own collection and actions.

Resolve <skill-dir> to this installed folder:

jev-decide decide <skill-dir>/assets/example.json --dry-run
# After approval, send the edited request with the selected provider:
jev-decide decide /path/to/request.json --provider openrouter

Workflow

1. Agree on categories with short inclusions/exclusions. Use independent Nouls when records can have several labels; use Choice for one queue.

2. Preserve original record IDs. For multiple records, name the exact record ID in every question or send one request per record; one Choice over an entire inbox is not per-message classification.

3. Collect text only from files or accounts the user authorized. Classify before writing tags, moving messages or sending replies.

4. Return a reviewable table: record ID, category, urgency, uncertainty and intended next consumer. Keep other/missing-evidence records visible.

5. Before bulk work, apply smoke_test above. Include near-miss categories and user-labeled examples; retest when criteria, models or context organization change.

Context and parallelism

Jev does not inherit the agent's history. Give every request sufficient context:

the user's categories and priority policy, each record's text and relevant thread,

product/account facts, and known missing evidence. A last-message fragment is not

enough when earlier messages change its meaning; omit unrelated history and secrets.

For bulk triage, batch independent category, escalation and urgency questions over

shared state instead of serial LLM calls. Name the record ID in every question.

Use bounded concurrency for independent requests, with stable IDs, rate limits and

a cost/time budget. The host schedules calls; the CLI has no parallel scheduler.

Questions cannot read other answers in the same request: gather any newly needed

account evidence before a dependent follow-up. Low latency is a reason to use Jev

for the judgment stage, not to skip quality checks or automate mailbox changes.

Make it yours

Replace the example's evidence, candidate IDs and criteria together. Preserve a

no-match route when the real task can fall outside the labels. Agree on how the

host or person consumes each answer before enabling any automatic effect.

Precedent

Related project or author example. Our workflow is an adaptation,

not that project's code, an automatic installer, or a reproduced benchmark.

OpenRouter request contract.

Examples

Mark sponsor segments in a video · Support queue routing · Urgency screening

More workflows and local templates (references/scenarios.md). Browse across

examples when designing a solution; follow the guides and sources that help.