# MindFort

> MindFort is an autonomous AI security platform. AI agents continuously pentest web apps, APIs, and infrastructure — finding and validating real vulnerabilities, then generating verified patches you can merge with a click.

- Agents run continuously on a cadence you set, probing targets like an attacker would, rather than point-in-time scanning.
- Each finding is validated to cut false positives, and ships with a contextual patch and a threat model explaining the fix.
- Assessment modes trade speed for depth: Balanced (200 credits, ~4 hrs, frequent/daily use), Deep (400 credits, ~6 hrs, weekly/bi-weekly), and Ultra (800 credits, ~8 hrs, pre-release and release gates). Task agents (10 credits) handle smaller, directed work like validating findings or triaging bug bounty reports, drawing from the same credit pool.
- Remediation spans code (PRs in GitHub), cloud config (AWS/Azure/GCP IAM, security groups), and issue tracking (Jira, Linear).
- For full technical and API documentation, see the separate docs site and its own llms.txt.

## Core pages

- [Home](https://www.mindfort.ai/): Overview of the platform — continuous AI pentesting, exploit validation, and automated remediation.
- [Product](https://www.mindfort.ai/product): How agents detect, validate, and remediate across code, cloud infrastructure, and network configurations.
- [About](https://www.mindfort.ai/about): Company mission, the case for autonomous security, and the founding team.
- [Support](https://www.mindfort.ai/support): Contact, demo booking, and answers to common product, pricing, and technical questions.

## Documentation

- [Docs home](https://docs.mindfort.ai): Setup, onboarding, assessments, findings, reporting, and integrations.
- [Docs index (llms.txt)](https://docs.mindfort.ai/llms.txt): Machine-readable index of all documentation, API reference, and MCP guides.
- [Quick Start](https://docs.mindfort.ai/quickstart): From account creation to your first completed assessment.
- [API Introduction](https://docs.mindfort.ai/api-reference/introduction): Programmatically trigger assessments, tasks, and findings via REST.
- [MCP Server](https://docs.mindfort.ai/guides/mcp): Connect Cursor, Claude Code, or Codex to MindFort findings.

## Optional

- [Y Combinator profile](https://www.ycombinator.com/companies/mindfort): Company background, team, and hiring (YC X25).