Security audit
Security checks for vulnerabilities and agentic risk
The skill appears to be a non-exfiltrating contest anti-cheat helper, but its identity-review purpose is paired with inconsistent inputs, overbroad declared tools, and weak privacy/retention guidance.
Review this before installing in a real contest workflow. It should not be treated as final adjudication, and operators should align the actual CLI inputs with the documented identity-review process, restrict tool permissions to what the script needs, and define privacy, retention, and human-review controls before storing or acting on participant identity records.
This file contains natural-language instructions exclusively in Chinese and explicitly points AI agents to follow those instructions. Under the stated policy, forcing a specific language without user opt-in is a locale/language policy violation unless clearly justified or optional.
The text states '中文文档见 SKILL.md', while the file overall presents the skill in Chinese and does not offer an explicit language or locale choice. This can violate language/locale policy because it effectively constrains users to Chinese documentation without opt-in or an alternative language path.
The trigger text is broad enough that an agent may invoke this skill in loosely related 'contest' or 'identity' contexts without clear boundaries. Because the skill handles anti-cheat and identity-verification decisions, overbroad activation can expose sensitive data to an unsuitable workflow and produce inappropriate security judgments outside the intended UIBC domain.
The skill describes processing contestant identity data and archiving verification records, but it does not warn users about handling sensitive personal data, retention, minimization, or consent. In practice this can lead operators or downstream agents to collect, store, or redistribute identity records unsafely, increasing privacy and compliance risk around PII.
The documentation says the skill requires event_id, participant_json, and policy_version, but the declared interface schema lists different parameters (author_id, env_hash, etc.). This mismatch can cause an agent or user to supply the wrong data, leading to incorrect anti-cheat decisions, bypass of intended checks, or logging of unrelated identity information. In a security-sensitive identity verification workflow, contradictory interfaces are operationally dangerous even without code execution.
The file’s docstrings, CLI descriptions, validation errors, warnings, notes, and advice are all hard-coded in Chinese. This imposes a specific language on all users, and the file does not provide any opt-in, locale selection, or explicit documentation that the tool is intentionally limited to a Chinese-speaking or region-specific context.
The changelog explicitly states “English first,” which indicates a preferred language policy in the documentation without offering user choice or opt-in. This can conflict with language/locale policy expectations when the skill is otherwise bilingual and not clearly region-restricted.
Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.
out_list.append({"tool": META["slug"], "input": d, "errors": [str(e)], "rc": 2, "aigc_mark": AIGC})
print(json.dumps(out_list, ensure_ascii=False, indent=2))
sys.exit(0 if allok else 2)
args = {k: getattr(ns, k) for k in argnames}
try:
res, rc = _run(args)
except GateError as e: