The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill consists entirely of instructional markdown and configuration metadata.
[PROMPT_INJECTION]: The instructions do not contain attempts to override system prompts or bypass safety filters. Instead, they reinforce skepticism and editorial rigor.
[DATA_EXPOSURE_EXFILTRATION]: No hardcoded credentials, sensitive file paths, or network exfiltration patterns were identified.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution, package installations, or dynamic script loading.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process datasets (e.g., housing data) which constitutes an attack surface. However, the skill lacks any high-privilege capabilities (like shell access or file writing) that could be exploited. The instructions include a strong emphasis on evidence auditing and source verification as a countermeasure.

data-journalist