The Agent Skills Directory

[PROMPT_INJECTION]: No malicious patterns or bypass attempts were detected. The skill contains instructions to follow a specific persona's voice and frameworks, which is the intended functionality. There are no attempts to override safety filters or extract system prompts.
[DATA_EXFILTRATION]: There is no evidence of sensitive data access or exfiltration. The skill does not contain hardcoded credentials, access sensitive file paths (like ~/.ssh or .env), or perform network operations to untrusted domains.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution or download external scripts. All referenced files are local markdown documents.
[OBFUSCATION]: No obfuscation techniques were identified. The content is written in plain, human-readable markdown without Base64 encoding, zero-width characters, or homoglyph attacks.
[INDIRECT_PROMPT_INJECTION]: The skill processes user input to select and load relevant educational topic files. While it interacts with user data, the ingestion is confined to local static files and lacks dangerous capabilities like subprocess execution or network writing that could be exploited via data poisoning.

dojo-paul-graham