jpmorgan-banker

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

• Third-party content exposure detected (high risk: 1.00). The skill's own required install instructions (e.g., "Read https://awesome-skills.dev/skills/enterprise/jpmorgan/jpmorgan-banker.md and install as skill" in Quick Install and Platform Support) instruct the agent to fetch and ingest content from a public third‑party URL, which could inject actionable instructions into the agent's prompts and behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs platforms to "Read https://awesome-skills.dev/skills/enterprise/jpmorgan/jpmorgan-banker.md" (including echoing that read command into system/config files), which indicates the remote markdown would be fetched at install/runtime and used to control the agent's system prompt—URL flagged: https://awesome-skills.dev/skills/enterprise/jpmorgan/jpmorgan-banker.md.