# AgentGG

- [Overview](https://docs.agentgg.dev/platform/overview.md): The agentgg platform runs the same agents as the CLI. It needs no install and provides the model for every scan.
- [Quickstart](https://docs.agentgg.dev/platform/quickstart.md): Sign in, create an organization, add credits, then start your first scan on the platform.
- [Organizations](https://docs.agentgg.dev/platform/organizations.md): Every scan and every finding belongs to an organization. Learn how to create one and switch between them.
- [Account](https://docs.agentgg.dev/platform/account.md): Change your name, connect or disconnect sign-in methods, and see your organization.
- [Start a scan](https://docs.agentgg.dev/platform/scans/start-a-scan.md): Scan a connected repository, a public repository URL, or a .zip archive you upload.
- [Connect GitHub](https://docs.agentgg.dev/platform/scans/connect-github.md): Install the AgentGG App, choose repositories, then turn on pull request scans for each one.
- [Manage a scan](https://docs.agentgg.dev/platform/scans/manage-a-scan.md): Watch a scan run, then handle one that paused, failed, stopped early, or was cancelled.
- [Findings](https://docs.agentgg.dev/platform/findings.md): The findings list, its filters, and the detail view for one finding.
- [Billing](https://docs.agentgg.dev/platform/billing.md): You buy credits up front, then each scan draws on them. You pay for the work a scan does, not a flat fee.
- [Troubleshooting](https://docs.agentgg.dev/platform/troubleshooting.md): Real messages from the upload card and the scan page, with the cause and the remedy for each.
- [Overview](https://docs.agentgg.dev/cli/overview.md): agentgg is a command-line security scanner. It reads your code and confirms a finding before it reports it.
- [Install](https://docs.agentgg.dev/cli/install.md): Install agentgg from npm, or build it from source. It requires Node.js 20 or later.
- [Quickstart](https://docs.agentgg.dev/cli/quickstart.md): Set up a provider, run a scan, then read the findings. Five steps from install to report.
- [Output and state](https://docs.agentgg.dev/cli/output-and-state.md): What a scan writes to its output directory, and why the state directory matters.
- [How a scan runs](https://docs.agentgg.dev/cli/how-a-scan-runs.md): The phases in order, how batches form, and how `--concurrency` bounds every phase from one pool.
- [Resume and reports](https://docs.agentgg.dev/cli/resume-and-reports.md): What makes a re-run reuse prior work, when to force a fresh pass, and the shape of a scan's reports.
- [Scan a pull request](https://docs.agentgg.dev/cli/guides/scan-a-pull-request.md): Restrict a scan to the files one commit or one range touched, with --diff.
- [Choose agents](https://docs.agentgg.dev/cli/guides/choose-agents.md): Restrict a scan to specific agents with -t, and see which categories run by default.
- [Validate findings and set scope](https://docs.agentgg.dev/cli/guides/validation-and-scope.md): Re-check each finding against the source, apply trust-boundary rules, then score and de-duplicate what survives.
- [Limit what gets scanned](https://docs.agentgg.dev/cli/guides/limit-what-gets-scanned.md): Narrow a scan to specific files, drop folders you don't own, and cap cost with per-agent and per-scan limits.
- [Plan a scan and inspect results](https://docs.agentgg.dev/cli/guides/plan-and-inspect.md): Preview which agents would run before you pay for a scan, then read or browse the results without another scan.
- [Run agentgg in GitHub Actions](https://docs.agentgg.dev/cli/guides/github-actions.md): Scan every pull request automatically, restricted to the files it changed, with one workflow file.
- [Providers](https://docs.agentgg.dev/cli/providers.md): Set up any of six model providers, or pass a one-off credential for a single scan.
- [Commands](https://docs.agentgg.dev/cli/reference/commands.md): Every agentgg command in one table.
- [Scan flags](https://docs.agentgg.dev/cli/reference/scan-flags.md): Every flag for agentgg scan, grouped by purpose, with its default.
- [Configuration](https://docs.agentgg.dev/cli/reference/configuration.md): Where agentgg stores its config, how to read it, and how it differs from scan state.
- [Troubleshooting](https://docs.agentgg.dev/cli/troubleshooting.md): Real messages from `agentgg scan` and its flags, with the cause and the remedy for each.
- [Overview](https://docs.agentgg.dev/agents/overview.md): An agent is one markdown file with frontmatter and a prompt. Learn the three parts, the categories, and where agents come from.
- [Agent anatomy](https://docs.agentgg.dev/agents/anatomy.md): Every agent is one markdown file. This page lists its frontmatter fields, the slug rule, and the noiseTier values.
- [Targeting](https://docs.agentgg.dev/agents/targeting.md): How an agent decides whether to run on a repository, and which files inside it to review.
- [Create from reports](https://docs.agentgg.dev/agents/create-from-reports.md): Turn a past security report into a reusable agent file with agentgg create.
- [Manage agents](https://docs.agentgg.dev/agents/manage.md): Install, inspect, and remove agents with agentgg agents, then contribute one back to the catalog.
- [Troubleshooting](https://docs.agentgg.dev/agents/troubleshooting.md): Real messages from `agentgg agents`, `agentgg create`, and the `-t` flag, with the cause and the remedy for each.