The Agent Skills Directory

[SAFE]: The skill uses local JSON and Markdown files (references/survey-free.json, references/user-survey-free.json, references/personality-types.md, references/scoring.md) to provide assessment logic and reference data. All data processing is contained within the agent's context with no external data exfiltration risks.
[SAFE]: No network operations, external downloads, or third-party dependencies are required for the skill's operation.
[SAFE]: The instructions and scripts do not attempt to execute shell commands, escalate privileges, or modify system files.
[SAFE]: The interaction model involves structured surveys which do not expose the agent to indirect prompt injection risks beyond standard conversational boundaries, and the skill lacks any high-risk capabilities (such as tool execution or file writing) that could be exploited.

agent-mbti