korean-spell-check

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly posts input to the public Nara/PNU web endpoint (see SKILL.md and DEFAULT_RESULTS_URL = "https://nara-speller.co.kr/old_speller/results") and parses untrusted HTML/payload fields (errInfo/help/errMsg) returned by that site to build corrections and reasons that directly affect outputs and behavior, so third‑party content can materially influence the agent.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill's runtime code posts user text to and parses responses from https://nara-speller.co.kr/old_speller/results, and those returned payloads are parsed and used to directly drive the agent's correction output, making this external URL a required runtime dependency that controls the agent's behavior.