pg-sell

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt instructs interactive collection of API keys and shows examples that embed secrets verbatim (CLI flags like --key and SDK examples with api_key: 'sk-new-key...'), so an agent could be required to output secret values directly, creating exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.80). The skill includes commands that browse and fetch public, user-generated marketplace content (e.g., "proxygate apis" and "proxygate listings docs " in SKILL.md and references/commands.md) so the agent can read external listing descriptions and docs which could influence subsequent actions.