Foo Guard
Menu

AI AGENT CONFIGURATION SCANNER

Find risky permissions
in your agent config.

Paste a JSON or YAML configuration. See which permissions, credentials, and approval settings need attention, with the evidence behind each finding.

Run a free scan See an example

No account required. The scanner reads the configuration without running your agent.

CONFIGS & SKILLS

Keep the version you reviewed.

Save configs and Markdown skills in Library, along with their versions and scan history. Your team can find the file and the review in one place.

  1. 01

    Add your file

    Paste or upload a configuration or skill to your Personal or Team Library.

  2. 02

    Review the findings

    Scan the saved copy. Earlier versions and their results stay available when you make changes.

  3. 03

    Publish a team config

    Publish an eligible configuration version after it passes your team’s policy. Teammates can retrieve that version from Library.

Published configs are sanitized copies; your team still chooses where to use them. Skills can be stored and scanned, but cannot be published.

See how Library works

01 / AGENT SECURITY SCANNING

See the setting behind each finding.

A finding shows the rule, severity, configuration field, and suggested change. Inspect the evidence before deciding whether to narrow a permission or change how the agent works.

Agent scan showing prioritized findings, the exact configuration evidence, and recommended remediation.
Scan findings and configuration evidenceView full size

02 / HOSTED GATEWAY

Review the requests your agents send.

Route supported MCP requests through Gateway to see allowed, flagged, and blocked activity. Team Enforcement can block requests that violate your policy. Calls that bypass Gateway are outside its coverage. Explore Gateway.

Gateway request history with blocked, allowed, and previously flagged tool calls, filters, and investigation actions.
Gateway Beta · Product UI with example dataView full size

03 / GITHUB & CI/CD

Review permission changes in the pull request.

GitHub checks show new, existing, and resolved findings against the pull request’s base revision. Set a policy threshold, then require the check in branch protection to gate merges. You can also run scans with the GitHub Action.

GitHub App repository scan showing failed policy, risk score 100, and no regression against the established baseline.
GitHub App scan · No regression does not mean low riskView full size

04 / AI-ASSISTED REMEDIATION

Check a suggested fix before you use it.

Generate a suggested configuration, review the changes, and check it from the remediation screen. Completed suggestions stay with your saved scan, so reopening them does not spend another AI credit. The scanner’s rules determine severity and scores.

AI Security Assistant explaining a missing human approval finding and presenting an advisory configuration change for review and re-analysis.
AI remediation · Product UI with illustrative guidanceView full size

Start with one configuration.

Paste your config or choose a sample to see how findings work.

PricingRun a free scan