Each agent can back several deployments (one per tenant/environment/runtime). A deployment
binds a pinned agent version to triggers and credentials.
jira-triagev1
Read a newly-created Jira ticket and draft a triage comment for a human to approve.
jira.issue.read, confluence.page.read, jira.comment.post
owner cloud-engineering
platform-supportv3
Investigate a platform support request and propose a response or PR review.
github.pull_request.read, github.review.post, slack.message.post
owner cloud-engineering
slack-channel-monitorv1
Read a message polled from a monitored Slack channel and draft a reply for a human to approve.
slack.thread.read, slack.message.post
owner cloud-engineering
slack-supportv1
Read a Slack support thread and draft a reply for a human to approve.
slack.thread.read, slack.message.post
owner cloud-engineering
tenant-context-testv2
Investigate the request using the available tools. Produce a concise summary, findings with evidence, and a proposed act
tenant.context.read
owner cloud-engineering
workload-healthv1
Watch one workload's health when an alert fires, correlate its traffic, resource, and log signals, and propose a proactive action (scale up before saturation, or rollout-restart on a known bad log pattern) within the tenant's action envelope. A human approves before anything happens.
tenant.context.read, metrics.query, k8s.workload.read, logs.query, k8s.workload.scale, k8s.pod.restart
owner cloud-engineering
| deployment | env | runtime | triggers |
|---|
| workload-health-k8s |
development |
k8s-job |
cli, grafana-alert, metric-threshold |