The Agent Skills Directory

[SAFE]: No prompt injection or behavior override patterns detected. The instructions strictly define a role for communication analysis without attempting to bypass safety filters or ignore prior constraints.
[SAFE]: No data exposure or exfiltration risks were identified. The skill does not reference sensitive file paths (e.g., .env, .ssh, .aws) and contains no network-capable commands like curl or wget.
[SAFE]: No obfuscation techniques were found. The content is written in clear, plain text (Chinese and English) without Base64 encoding, zero-width characters, or hidden unicode patterns.
[SAFE]: The skill does not include any external dependencies, package installations, or remote code execution patterns. It operates purely as a set of logical instructions for the AI agent.
[SAFE]: There are no signs of privilege escalation or persistence mechanisms. The skill does not attempt to modify system files, startup scripts, or acquire administrative rights.
[SAFE]: Indirect prompt injection risk is negligible. While the skill ingests user content for analysis, it lacks the capabilities (such as file-writing, network access, or code execution) that would make such an injection exploitable. The output is constrained to a predefined analytical template.

dbs-spread