# Ace integrations: connect your tools, connect the evidence

> Give Ace's agents access to the tools your engineers already use. They follow the failure across infrastructure, telemetry and code, and share findings in Slack.

## Infrastructure

- **AWS**: Resource relationships and service context across connected AWS accounts, through a read-only role that denies secret values and decryption. (Resource topology · account & region context)
- **Kubernetes**: An outbound connector for EKS, GKE, AKS and on-premises clusters, installed with Helm. Namespace-scoped reads, and a separate identity you grant before Ace can restart, scale or patch anything. (Resources · events · logs · rollouts)
- **PlanetScale**: Read schema, branches and row counts. Ace proposes schema changes as deploy requests and bounded backfills, and a person always approves them. (Schema · branches · deploy requests)
- **CloudTrail**: Check recent AWS changes, who made them and when, against the incident timeline. (Change events · actors)

## Code & delivery

- **GitHub**: Connect code and deployments to investigations. Ace opens fix and revert pull requests, fixes its own failing checks once, and can run a rollback workflow your administrator confirms. (Repositories · deployments · pull requests)
- **Vercel**: Correlate deployments with incidents. Ace can propose an instant rollback to a previous production deployment, or a promotion. (Deployments · promotions)
- **PostHog**: Measure which users an incident affects, and roll a feature flag out in stages with checks between each one. (Product impact · feature flags)
- **Trigger.dev**: Failed background runs are grouped into incidents. Ace can replay failed runs in stages, each run at most once. (Runs · failures · replays)

## Observability

- **Datadog**: Query APM traces, metrics and monitors from your Datadog account during an investigation. (APM · metrics · monitors)
- **Sentry**: Errors, issues and releases. Issue alerts open an investigation, and repeated errors are grouped into one incident. (Issues · events · releases)
- **Prometheus**: Query metrics from Amazon Managed Prometheus and EKS setups. Alertmanager alerts can open an investigation. (Time series · workload signals)
- **Grafana**: Grafana alerts open an investigation. Firing alerts and alert rules inform the evidence. (Alerts · alert rules)
- **Elasticsearch**: Search connected logs for evidence within the incident window. (Logs · scoped queries)
- **Axiom**: Query logs and events in the datasets you choose, with bounded APL over the incident window. (Logs · events)
- **LangSmith**: Check LLM application traces alongside the rest of the incident evidence. (LLM traces · application context)
- **Logfire**: Read application telemetry from Logfire during an investigation. (Application telemetry)
- **LiteLLM**: LLM gateway health, model errors and spend. Scheduled scans look for duplicate and wasted model spend. (Gateway health · model errors · spend)
- **CloudWatch**: Bring alarms, logs and metrics into an investigation. Alarm notifications sent through Amazon SNS can open one. (Metrics · logs · alarms)
- **AWS X-Ray**: Follow requests across instrumented services and dependencies. (Traces · service relationships)

## Collaboration

- **Slack**: Follow the investigation in the incident thread. Approve, decline or ask about a proposed action there, and see each step as it runs. (Updates · approvals · action progress)

## Coming next

- **Azure**: Azure infrastructure and service dependencies.
- **Supabase**: Database and application infrastructure context.
- **MS Teams**: Incident updates in Microsoft Teams.

## Access

Each connection has its own permissions and scope, and investigation is read-only. See [security](https://tryace.dev/security) for what each connection can read. We check source coverage and required permissions with your team. Missing a tool? Email [founders@vratix.com](mailto:founders@vratix.com) with what you use and what Ace needs to read.