> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cloudeval.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Capabilities map
> See which CloudEval AI capabilities are current, in progress, planned, or limited.
Use this page when you need a quick, honest map of what CloudEval can do today and where the product is headed next.
Roadmap items are not shipped behavior. Treat anything marked `In progress` or `Planned` as directional until the docs or `cloudeval capabilities --format json` say it is current.
## Why it matters
Cloud evaluation tools are only useful when teams know what they can trust. This map separates current product capability from roadmap direction so architects, DevOps teams, and AI agents can plan without guessing.
## Quick check
For automation, check the live CLI contract first:
```bash theme={null}
cloudeval capabilities --format json
cloudeval mcp status --format json
cloudeval doctor --format json
```
Expected result:
```json theme={null}
{
"ok": true,
"command": "capabilities",
"data": {
"commands": [],
"domains": {},
"mcp": {}
}
}
```
The real output includes the current command list, frontend deeplinks, MCP toolsets, resources, and prompts.
## Status labels
Current
In progress
Planned
Limited
| Status | Meaning |
| ----------- | -------------------------------------------------------------------------------------------------- |
| Current | Available in the product or CLI today and safe to document as a supported workflow. |
| In progress | Active roadmap work, not generally available yet. Use it for planning, not production commitments. |
| Planned | Directional roadmap item. Scope and dates can change. |
| Limited | Available only for specific workflows, preview paths, or narrow usage. |
## Capabilities
| Area | Capability | Status | What users get | Start here |
| -------------------- | --------------------------------------------------------------------- | ------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- |
| Cloud ingestion | Azure Cloud sync connections | Current | Connect an Azure subscription or resource group scope and turn it into a CloudEval project. | [Connect Azure](/quickstart/connect-an-azure-environment) |
| IaC ingestion | Single ARM template import | Current | Upload or reference ARM JSON before deployment and evaluate it as a project. | [Import ARM or Bicep](/quickstart/import-an-arm-or-bicep-template) |
| IaC ingestion | Multi-file ARM workspace import | Current | Preserve linked or nested ARM templates as source files and resolve them for diagrams and reports with `.cloudeval/config.yaml`. | [IaC project config](/reference/iac-project-config) |
| IaC ingestion | Bicep review through compiled ARM JSON | Current | Compile Bicep to ARM JSON, then use the same import and report workflow. | [Automate with the CLI](/workflows/automate-evaluations-with-the-cli) |
| IaC ingestion | GitHub source snapshot trust model | Current | Understand how repository files are imported, stored, filtered, and used for reports and AI-assisted review. | [Data and model boundaries](/reference/data-and-model-boundaries) |
| Project model | Projects, connections, files, and sharing | Current | Keep topology, source inputs, reports, collaboration, and share links together. | [Projects](/concepts/projects) |
| Diagrams | Interactive architecture and dependency diagrams | Current | Inspect cloud resources, relationships, layout, and resource context visually. | [Diagrams and workspace views](/concepts/diagrams) |
| Reports | Cloud Evaluation report | Current | Review Brief, Action Plan, Architecture & Dependencies, Security & Compliance, Cost & FinOps, Resources, and Evidence & Progress in one project-aware report. | [Reports](/concepts/reports) |
| Reports | Cost and FinOps reporting | Current | Estimate run-rate cost, annualized context, savings opportunities, service-family breakdowns, pricing assumptions, and unmatched resources. | [Run reports](/quickstart/run-your-first-reports) |
| Reports | Architecture and Well-Architected reporting | Current | Review pillar scores, findings, issue counts, dependency context, affected resources, and architecture recommendations. | [Report types](/reference/report-types-and-statuses) |
| Reports | Evaluation coverage | Current | Understand the `1,700+` CloudEval platform signal floor, `1,650+` attributed third-party catalog, workflow overlap, and applicability limits. | [Evaluation coverage](/reference/evaluation-coverage) |
| Reports | Resource inventory and finding detail | Current | Triage exact affected resources with status, severity, issue count, cost, savings, evidence links, and primary action. | [Reports](/concepts/reports#resources-and-evidence) |
| Reports | Cross-project issues inventory | Current | Triage architecture findings, cost opportunities, and unit-test failures across projects in `/app/issues` or with `cloudeval issues list`. | [CLI command reference](/reference/cli-command-reference#issues-inventory) |
| Reports | Security and architecture finding review | Current | Triage risk with report evidence, confidence, source context, and project links. | [Review findings](/workflows/review-cost-and-architecture-findings) |
| Reports | Report exports and AI handoff | Current | Export PDF, Markdown, or JSON and copy scoped report context or evidence packs for agent-assisted follow-up. | [LLM context](/reference/llms-and-agent-context) |
| Reports | ARM/unit-test reports where available | Current | Run template-focused checks when the project and CloudEval service support them. | [Report types](/reference/report-types-and-statuses) |
| AI workspace | Grounded project chat | Current | Ask questions against project metadata, graph data, source inputs, and saved reports. | [Agent rules](/reference/agent-and-automation-rules) |
| CLI | Agentic CLI workflows | Current | Configure profiles, create projects, run reports, ask grounded questions, download outputs, and open deeplinks. | [CLI overview](/reference/cli-overview) |
| Agent integration | MCP server | Current | Expose CloudEval project, report, billing, open-url, and capability tools to MCP-compatible agents. | [Agent rules](/reference/agent-and-automation-rules) |
| Agent integration | `llms.txt` and `llms-full.txt` | Current | Give external agents machine-readable product, CLI, grounding, and limitation context. | [LLM context](/reference/llms-and-agent-context) |
| Collaboration | Share links, embeds, and invited roles | Current | Share read-only project views, embed snippets, or collaborate with named users. | [Sharing](/concepts/sharing-and-collaboration) |
| Billing | Credits, Pro checkout, plan visibility, and CLI billing summaries | Current | Inspect usage, credits, the enabled Day Pass and Pro plans, ledger entries, invoices, top-ups, and billing deeplinks. | [Plans and billing](/plans-and-pricing) |
| Pull requests | GitHub App repository sync | Current | Link a GitHub repository to a CloudEval project so review jobs can resolve repository, branch, commit, and source-root provenance. | [GitHub repository sync](/reference/github-repository-sync) |
| Pull requests | GitHub Actions composite action | Current | Run CloudEval review in CI, upload workflow artifacts, and use a scoped access key instead of hand-rolled CLI setup. | [GitHub Actions integration](/workflows/github-actions) |
| Pull requests | PR comments from CloudEval review | Current | Post or update one result comment with source provenance, report links, gate result, and drilldowns when repository permissions allow it. | [GitHub Actions integration](/workflows/github-actions#pr-review-workflow) |
| Pull requests | CI job failure via configured gates | Current | Fail the workflow job when `.cloudeval/config.yaml` `ci.gates` thresholds are not satisfied. | [GitHub Actions integration](/workflows/github-actions#config-driven-gates) |
| Editor workflow | VS Code sidebar for CloudEval projects and findings | In progress | Review projects, resources, reports, and findings without leaving the editor. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Editor workflow | Inline ARM/Bicep diagnostics and CodeLens actions | In progress | See IaC risks in context, then explain, evaluate, or draft fixes from the editor. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Editor workflow | CloudEval MCP tools for VS Code agents | In progress | Let editor agents answer with CloudEval project, graph, and report context. | [Agent rules](/reference/agent-and-automation-rules) |
| Pull requests | GitHub Checks UI annotations | Planned | Add richer GitHub-native annotations beyond the workflow job result and PR comment. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Pull requests | PR diagram diffs | Planned | Show architecture diffs across pull-request changes. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Pull requests | Native repo scanning without setup | Planned | Scan repositories without requiring the current GitHub App plus action workflow setup path. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Diagnostics | Snapshot drift and graph diff | Planned | Compare project snapshots to see what changed and which risks moved. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Diagnostics | Blast radius and risk-path exploration | Planned | Trace dependencies, identities, network paths, exposure, data access, and cost impact. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Provider coverage | AWS CloudFormation scanner evaluation | Limited | Static beta evaluation for CloudFormation using reviewed AWS beta controls, deployment-quality checks, and additional IaC findings where applicable. | [Cloud support](/cloud-support) |
| Provider coverage | Terraform scanner-native findings | Limited | Additional IaC findings for supported Terraform workflows. Do not treat this as full Well-Architected or live-cloud parity. | [Cloud support](/cloud-support) |
| Provider coverage | GCP, Kubernetes, Pulumi, and telemetry evidence | Planned | Broader coverage with explicit parser, provider, rule, and evidence labels. | [Cloud support](/cloud-support) |
| Remediation | Ranked fixes, IaC patches, PRs, waivers, owners, and evidence exports | Planned | Turn findings into reviewable change proposals with governance context. | [Roadmap](https://cloudeval.ai/home/roadmap) |
| Trust and deployment | Self-hosted or private-network deployment | Planned | Keep source snapshots, cloud sync evidence, and generated reports inside customer-controlled infrastructure for Team and Enterprise use cases. | [Data boundaries](/reference/data-and-model-boundaries) |
| Trust and deployment | Custom LLM gateway and bring-your-own key | Planned | Route AI summaries, chat, and review context through an approved enterprise model gateway or customer-owned model account. | [Data boundaries](/reference/data-and-model-boundaries) |
| Provider coverage | AWS and GCP live connection parity | Limited | Do not assume full Azure parity today. AWS live connections and GCP live connections are not available today. | [Cloud support](/cloud-support) |
| Commerce | Team, Business, and enterprise self-serve checkout | Limited | Team and Business are not available for in-app purchase today; Enterprise remains contact sales. | [Plans and billing](/plans-and-pricing) |
## How to use this map
1. Start with `Current` rows when planning a real rollout.
2. Use `In progress` rows for near-term pilots and design discussions.
3. Use `Planned` rows to understand direction, not to make delivery commitments.
4. Check `Limited` rows before assuming provider, billing, or integration parity.
## Common mistakes
* Treating roadmap rows as available product behavior.
* Treating planned Team/Enterprise self-hosting or custom LLM routing as available in the public Pro product.
* Assuming AWS, GCP, Terraform, CloudFormation, or Kubernetes support matches Azure live-cloud support.
* Adding workflow scanner counts together.
* Hard-coding CLI commands without first running `cloudeval capabilities --format json`.
* Registering MCP broadly when a focused toolset such as `readonly`, `projects`, `reports`, or `billing` is enough.
## Tips
* Use [Feature matrix](/feature-availability) for the strict support matrix.
* Use [Cloud and IaC support](/cloud-support) before writing provider-specific runbooks.
* Use [Evaluation methodology and attribution](/reference/evaluation-methodology-and-attribution) before quoting counts, pinned versions, or third-party attribution.
* Use [Agent and automation rules](/reference/agent-and-automation-rules) before wiring CloudEval into an AI agent.
* Use [Data and model boundaries](/reference/data-and-model-boundaries) before connecting private GitHub repositories or planning enterprise model-routing requirements.
* Use the [public roadmap](https://cloudeval.ai/home/roadmap) when you need the six-month product direction.
## Related pages
* [Feature matrix](/feature-availability)
* [Evaluation coverage](/reference/evaluation-coverage)
* [Evaluation methodology and attribution](/reference/evaluation-methodology-and-attribution)
* [CLI command reference](/reference/cli-command-reference)
* [Agent and automation rules](/reference/agent-and-automation-rules)
* [Data and model boundaries](/reference/data-and-model-boundaries)
* [Cloud and IaC support](/cloud-support)
* [Plans and billing](/plans-and-pricing)