# Documentation

## Docs

- [Agent Security](https://atcyrus.com/docs/agent-security.md): How Cyrus protects your organization from prompt injection, tool abuse, and autonomous agent risks.
- [Allowed Tools per Platform](https://atcyrus.com/docs/allowed-tools.md): Configure which tools Cyrus can use, separately for Linear, Slack, and GitHub
- [Connecting Your Tools](https://atcyrus.com/docs/connecting-tools.md)
- [Deploy Previews](https://atcyrus.com/docs/deploy-previews.md): How to use Vercel deploy previews to review Cyrus's changes end-to-end before merging.
- [Create First Issue](https://atcyrus.com/docs/first-issue.md)
- [Installation](https://atcyrus.com/docs/getting-started.md)
- [Git Committer Identity](https://atcyrus.com/docs/git-committer-identity.md): How Cyrus configures the Git committer identity for agent-made commits.
- [How Cyrus Chooses Work](https://atcyrus.com/docs/how-cyrus-knows.md)
- [GitHub](https://atcyrus.com/docs/integrations/github.md): How Cyrus integrates with GitHub — PR mentions, review triggers, and more
- [Linear](https://atcyrus.com/docs/integrations/linear.md): How Cyrus integrates with Linear: assigned issues, label triggers, and comment mentions
- [Custom MCPs](https://atcyrus.com/docs/integrations/mcps.md): Add and manage custom MCP servers from the Cyrus dashboard using API key or token-based authentication.
- [Slack](https://atcyrus.com/docs/integrations/slack.md): How Cyrus integrates with Slack — @mention triggers, thread context, and MCP tools
- [Cyrus](https://atcyrus.com/docs/introduction.md)
- [Labels and Routing](https://atcyrus.com/docs/labels-and-routing.md)
- [Monitoring](https://atcyrus.com/docs/local-setup-post.md)
- [Prerequisites](https://atcyrus.com/docs/local-setup-pre.md)
- [Managing Repositories](https://atcyrus.com/docs/managing-repos.md)
- [MCP Servers](https://atcyrus.com/docs/mcp-servers.md): Connect custom MCP servers (like Stripe) so Cyrus can safely interact with external systems.
- [Anthropic (Claude)](https://atcyrus.com/docs/providers/anthropic.md): How to configure Anthropic (Claude) for use with Cyrus
- [Cursor](https://atcyrus.com/docs/providers/cursor.md): How to configure Cursor for use with Cyrus
- [Default Model Provider](https://atcyrus.com/docs/providers/default-model-provider.md): How to change the default model provider used for every issue and agent run
- [OpenAI](https://atcyrus.com/docs/providers/openai.md): How to configure OpenAI for use with Cyrus
- [Runtimes](https://atcyrus.com/docs/runtimes.md)
- [Setup scripts](https://atcyrus.com/docs/setup-scripts.md)
- [Skills](https://atcyrus.com/docs/skills.md): Teach Cyrus your team's playbooks (review checklists, deploy runbooks, triage flows) with portable, scoped capabilities the agent picks up at runtime.
- [Troubleshooting](https://atcyrus.com/docs/troubleshooting.md)
- [What Cyrus Manages](https://atcyrus.com/docs/what-cyrus-manages.md)
- [Writing Great Issues](https://atcyrus.com/docs/writing-great-issues.md): The best way to get strong hands-off results from Cyrus, and other AI, is to provide a strong 'definition of done', also known as 'acceptance criteria'. Here's a tip for drafting issues.

## OpenAPI Specs

- [openapi](https://atcyrus.com/docs/api-reference/openapi.json)