pp-yeswehack

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.78). SKILL.md describes an agent workflow where the CLI is invoked against YesWeHack and can output “hacktivity”/disclosed reports, which are outsider-authored free text from public/private platform feeds, likely routed into the agent/LLM via the --agent JSON output (e.g., hacktivity learn, programs hacktivity, report dedupe).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill directs a runtime installation using "go install github.com/mvanhorn/printing-press-library/library/developer-tools/yeswehack/cmd/yeswehack-pp-cli@latest", which fetches and builds remote Go code that would execute as a required dependency for the CLI.