agent-incident

Triage-shaped, not admin-shaped — the commands incident response needs, not 165+ incident.io API endpoints

Compact by default — an incident comes back as id, name, status, severity, created_at and lead until you ask for --full

Flexible incident ids — INC-2000, 2000 or a UUID all work, and get takes several at once

Item-level misses do not lose the batch — an unresolvable id becomes an @unresolved line on stdout with exit 0; only auth and network failures exit non-zero

Errors classify themselves — fixable_by: agent means fix your parameters, human means stop and ask, retry means wait

Cursor pagination handled automatically, so a list is a list rather than a paging exercise

On-call in full — schedules, who is on now, entries over a window, overrides, escalation paths and creating an escalation

Status pages and their updates, for the part of an incident customers see

Reference data alongside — severities, statuses, roles, users, custom fields, timestamps and the catalog

Keys never touch argv — --form opens a native OS dialog, or pipe the key on stdin; INCIDENT_API_KEY also works with no setup

Multi-org — several incident.io organizations side by side, selected with --org

Self-documenting — agent-incident usage, and a usage page per command group

Single compiled Go binary — no runtime dependencies