Devops Watch Stack
devops_watch_stackIdempotent
Check the health of a named vendor stack — a saved list of vendors representing your infrastructure dependencies. On the first call, provide vendors to define the stack; subsequent calls can omit vendors to reuse the saved list. Stacks are saved per authenticated caller, or per local server process; a shared deployment that does not authenticate callers saves nothing, so there vendors is required on every call and the response says so. Up to 20 vendors are saved per stack and checked per call; use several stack names for a larger set. Returns a unified health snapshot with an aggregate rollup plus per-vendor detail. A vendor that cannot be resolved or reached is reported in its own row and left out of the saved stack, so one bad entry never discards the sweep. Every vendor with an active problem gets a pre-filled devops_suggest_action call in nextToolSuggestions. Ideal for morning status checks or pre-deploy sweeps. Multiple stacks can coexist (e.g., "production", "staging").
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| mode | No | summary: indicator + degraded components + active incidents. detailed: adds component lists and maintenance windows. | summary |
| vendors | No | Vendor slugs (e.g., "github", "aws") or raw Atlassian Statuspage base URLs. When provided, saves this list as the stack where stacks are saved; a list longer than 20 is then rejected and nothing is saved. When omitted, uses the previously saved list for stack_name; required on every call on a deployment that does not authenticate callers, which saves no stacks — there the first 20 are checked and the rest return in deferred_vendors. Lists longer than 100 are rejected. | |
| stack_name | No | Name for this vendor stack. Defaults to "default". Use distinct names to manage multiple stacks (e.g., "production", "data-layer"). Letters, digits, hyphens, and underscores, optionally separated by single dots or slashes ("prod.eu", "team/prod"); 1-64 characters. No spaces or colons. | default |
| component_limit | No | Maximum components returned per vendor in detailed mode, at least 1. A value above 500 is reduced to 500, and notice says so. Large status pages publish hundreds of components, so a full stack at a high limit returns a very large response; narrow with component_filter instead where possible. | |
| component_filter | No | Case-insensitive substring matched against component names and their group names in detailed mode (e.g., "api" to check just the API components, or a product or region a page groups components under). Applied before component_limit, so it is the way to reach a component that the cap would otherwise omit. A filter that matches nothing for a vendor is reported in notice. Ignored in summary mode. |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| cap | No | The per-vendor component_limit that was applied. Present only when truncated. | |
| error | No | Present when the call failed. Absent on success. | |
| shown | No | Components returned across the stack. Present only when truncated. | |
| health | No | Aggregate health rollup: all_operational = everything clear, maintenance = at least one vendor in a scheduled window and nothing worse open, degraded = at least one minor issue, partial_outage = at least one major issue, major_outage = at least one critical outage, unknown = at least one vendor could not be checked (unresolvable entry, blocked target, or failed fetch) and no checked vendor reported a worse issue. Never all_operational when any vendor errored or is in a window. | |
| notice | No | Plain-language explanation of what this call left out or adjusted: vendors past the first 20 (listed in deferred_vendors), a component_limit above 500 reduced to 500, and in detailed mode how many components the cap omitted and how to reach them (component_filter to target one, or component_limit to raise a cap below 500), and which checked vendors component_filter matched no component of, with how many each publishes. Present only when one of these applies. | |
| summary | No | Aggregate health counts across all checked vendors. Buckets partition the stack: operational + degraded + down + maintenance + unavailable = total. | |
| vendors | No | Per-vendor status results for the vendors checked, in stack order. | |
| truncated | No | True when at least one vendor's component list was capped at component_limit. Absent when nothing was capped. | |
| checked_at | No | ISO 8601 UTC timestamp of this check. | |
| stack_name | No | Name of the stack checked. | |
| totalCount | No | Components matching component_filter across the stack before the cap. Present only when truncated. | |
| omitted_vendors | No | Entries that could not be resolved or whose URL was blocked; they still appear in vendors[] with an error. A call that saved the stack left them out of the write; a call that reused a saved stack leaves them in it until you re-provide the vendors list. Empty when every entry resolved. | |
| stack_persisted | No | True when the vendor list was saved on this call. Only the vendors that resolved are saved — see omitted_vendors. Always false on a deployment that does not authenticate callers, which saves no stacks (persistence_notice says so). | |
| deferred_vendors | No | Entries past the first 20, verbatim and in order, that were not checked on this call: the rest of a passed vendors list on a deployment that saves no stacks, or of a saved stack holding more than 20. Resend them in another call. Empty when every entry was checked. | |
| persistence_notice | No | Present when this deployment saves no stacks because callers are not authenticated: nothing was saved, and vendors must be passed on every call. | |
| nextToolSuggestions | No | One devops_suggest_action call per vendor with an active problem — an indicator of minor, major, or critical, or an open incident of that impact even while the indicator reads none or maintenance — with its arguments filled from that vendor's result. Vendors that could not be checked never appear. Empty when no checked vendor has an active problem. |