Security audit
Security checks for vulnerabilities and agentic risk
This skill is not clearly malicious, but it gives a broad persona workflow persistent influence over future agent behavior and local learning files without tight user control.
Install only if you want this skill to act as a broad, persistent persona and learning system, not just a temporary analysis style. Review and consider disabling the ~/.codex learning writes, broad triggers, forced identity rules, and automatic/background maintenance workflows before using it in normal Codex sessions.
The skill instructs writing to a home-directory path outside the project workspace (~/.codex/lessons/LESSONS.md) without prominently warning the user. Writing outside the workspace is more dangerous because it affects persistent global state, may bypass expected repository review, and can contaminate future agent behavior across projects.
The trigger list includes broad, everyday phrases such as '全面分析' and '帮我决策', which can cause the skill to activate in many unrelated conversations. In this skill, accidental activation is more concerning because it imposes a strong persona, hidden file-reading workflow, and cross-domain decision framing that may override user expectations and route normal requests into an unintended analysis mode.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
The skill is presented as a reasoning/persona aid, but the specification embeds many operational behaviors such as file reads/writes, report generation, routing maintenance, testing, audits, and script execution. This mismatch is dangerous because users and orchestrators may invoke a seemingly harmless analytical skill that actually performs repository or local-environment actions outside the declared purpose.
No suspicious patterns detected.
