Skip to content

Security: kenwalger/managed-agents-workshop

Security

SECURITY.md

Security policy

Scope

This repository is an educational workshop, not a production agent service. The exercises can create billable cloud resources and execute code in a managed sandbox. Use only synthetic fixture data and a dedicated development API key with appropriate account-level controls.

Reporting a vulnerability

Please do not open a public issue containing credentials, exploit details, or sensitive data. Use GitHub's Report a vulnerability / private vulnerability reporting feature if it is enabled for this repository. Otherwise, contact the repository maintainer privately through the contact information on their GitHub profile. Do not assume a guaranteed response time.

Include a minimal reproduction, affected files or versions, impact, and recommended mitigation if known. Never include live API keys or private user data.

Safe workshop operation

  • Keep ANTHROPIC_API_KEY in .env or your environment; never commit .env, logs containing secrets, or session state.
  • Confirm .env.example contains placeholders only.
  • Use the smallest tool permissions and network egress needed for the exercise.
  • Use a session spend cap; check account-level limits independently.
  • Review generated commands and files before executing or applying them locally.
  • Delete unused sessions and uploaded files, and archive agents when appropriate.
  • Treat agent-generated reports as untrusted data; verify against the provided fixture.
  • check.py follows this rule: it reproduces evidence only by running samples it parses from the local fixture/ folder, and it never runs the command fields or any other code from the agent's report. check.py stretch reads the agent's corrected docs without running them, and replay.py only prints recorded events.

Supported versions

Only the current default branch is maintained. This workshop is a learning artifact and does not promise security fixes or long-term support for earlier revisions.

There aren't any published security advisories