# The curated library

> The fifteen checkups shipped with the app, what each one looks at, and which ones need an integration connected first.

Source: https://triagic.com/docs/checkups/library

The library is grouped into seven categories, and the Checkups page renders one rail
per category with your own checkups pinned above them.

| Category | What it is for |
| --- | --- |
| **Optimize** | Money and performance in a system you pay for |
| **Review** | A periodic examination someone has to sign off on |
| **Hygiene** | Things that rot quietly: credentials, stray personal data |
| **Support quality** | How your team is actually serving customers |
| **Understand** | Documentation exercises: what you have, what your tickets are telling you |
| **Readiness** | Whether you could handle the thing that has not happened yet |
| **Triage** | Run against one named incident rather than on a cadence |

## Optimize [#optimize]

| Checkup | What it looks at | Needs |
| --- | --- | --- |
| **Snowflake cost and warehouse efficiency** | Idle and oversized warehouses, auto-suspend settings, expensive query patterns, storage nobody meant to keep. | Snowflake |
| **DynamoDB capacity and table provisioning** | Over- and under-provisioned tables, on-demand versus provisioned fit, unused GSIs, tables missing TTL. | DynamoDB |
| **Postgres performance and index health** | Slow query patterns, missing and unused indexes, table bloat, connection posture. | Postgres |

## Review [#review]

| Checkup | What it looks at | Needs |
| --- | --- | --- |
| **SOC 2 readiness review** | A Trust Services Criteria common-criteria walkthrough over your connected systems and support history, with evidence recorded per control. See [Compliance checkups](/docs/checkups/compliance) before you run it. | None |
| **Access review** | Who has access to what across your connected systems: dormant accounts, over-privilege, and shared credentials. | None |
| **GitHub code review** | An expert review of recently merged pull requests and default-branch commits: bugs, security issues, leaked secrets, and review-process gaps like unreviewed merges. | GitHub |
| **GitLab code review** | The same review over merge requests, for organizations on GitLab. | GitLab |

## Hygiene [#hygiene]

| Checkup | What it looks at |
| --- | --- |
| **Credential and integration hygiene** | Credential age and rotation, stale or failing connections, orphaned servers, integrations nobody uses any more. |
| **PII exposure sweep** | Personal data sitting in tickets, chat threads and reachable data stores: where it is, how much, and what should be redacted. |

## Support quality [#support-quality]

| Checkup | What it looks at |
| --- | --- |
| **SLA and response quality** | Breached targets, slow first responses, tickets left to rot, and reply-quality outliers, from your own ticket history. |
| **Knowledge and playbook gaps** | Recurring ticket clusters no playbook covers, plus drafted [triage instructions](/docs/playbooks/writing-instructions) for the largest gaps. |

## Understand [#understand]

| Checkup | What it looks at |
| --- | --- |
| **Environment map** | What is connected, what the agent can actually see and do with it, where it is used, and what is missing. |
| **Ticket trends and recurring root causes** | What volume is doing, the anatomy of each spike, and the root causes that keep coming back across months. |

## Readiness [#readiness]

| Checkup | What it looks at |
| --- | --- |
| **Incident readiness and runbook coverage** | Whether your playbooks and runbooks cover the incidents you actually have, and what an on-call responder could reach at 3am. |

## Triage [#triage]

| Checkup | What it looks at |
| --- | --- |
| **Webhook and integration failure deep-dive** | Run this one against a named ticket or incident: trace a webhook or integration failure end to end and find where events are being lost. Give it the ticket in the run's [focus](/docs/checkups/runs-and-findings#running-one-now). |

## What you can run on day one [#what-you-can-run-on-day-one]

Only five checkups require an integration: **Snowflake cost** needs Snowflake,
**DynamoDB provisioning** needs DynamoDB, **Postgres performance** needs Postgres,
and the two **code review** checkups need GitHub or GitLab respectively. A card
whose required integration is missing offers *Connect &lt;provider&gt;* instead
of *Run*.

The other ten run with nothing connected at all, because your own support data
(the ticket history, the investigations, the integration inventory) is already
enough to say something useful. That is deliberate: a new organization should not
meet a wall of locked cards. (The **access review** is the honest exception. With
no enumerable system connected it reports that it had nothing to review rather
than inventing a surface.)

> **Note:** A good first run
>
> **Environment map.** It needs nothing, it produces a document rather than a
> complaint, and its findings are mostly of the shape *here is a system you depend on
> that Triagic cannot see*, which tells you what to connect next.

Where an integration would make a checkup *richer* rather than possible, the
procedure says so and works without it. Connecting more systems generally makes
every checkup better, not just the three that insist.
