The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains only educational guidelines on software architecture and design principles. No instructions attempting to bypass safety filters or override agent behavior were detected.- [DATA_EXFILTRATION]: The analysis found no network-based commands (like curl or wget) or attempts to access sensitive system files or credentials.- [OBFUSCATION]: All content is in clear text. No Base64, hex encoding, zero-width characters, or hidden steganographic patterns were identified.- [REMOTE_CODE_EXECUTION]: The skill does not include any commands for installing external packages or executing remote scripts.- [PRIVILEGE_ESCALATION]: No commands for elevating user privileges (such as sudo or chmod) were found.- [PERSISTENCE]: There are no instructions aimed at maintaining long-term access, such as modifying shell profiles or creating scheduled tasks.- [METADATA_POISONING]: The name and description fields accurately represent the skill's content without deceptive or malicious directives.- [INDIRECT_PROMPT_INJECTION]: The skill consists of static markdown documentation and does not ingest or process untrusted external data that could lead to indirect injection.- [TIME_DELAYED_CONDITIONAL]: No logic based on date, time, or environment triggers was found.- [DYNAMIC_EXECUTION]: The skill does not generate or execute any code at runtime.- [DYNAMIC_CONTEXT_INJECTION]: No use of shell command execution placeholders (such as the !command syntax) was detected.

keystone