Code quality

Trace dangerous agent actions

/agent-exploit-paths

Find dangerous paths from untrusted input to privileged actions.

Put it to work

Check whether text from a web page could cause an agent to run an unwanted command.

Illustration: a red signal travels from an untrusted paper message toward a privileged brass key, stopped at an inspection gate

Better together

Give the next step a head start.

Suggested ways to combine this skill with the pack. Choose the handoff that fits your project.

Useful work before this

Find risky access paths

/security-sweep

For agent features, inspect paths from untrusted text to privileged actions.

Explore the whole connections map →

Get this skill—and the other 38.

The complete instructions and supporting files come in the BilbroSwagginz skill pack. Install them in Claude Code and use the skills across your projects.

One-time purchase. Requires Claude Code. Instant download.

Secure checkout via Stripe Instant download