The Agent Skills Directory

[SAFE]: The skill architecture is based on providing grounded religious teaching using provided text excerpts and external references to Buddhist sutras.
[COMMAND_EXECUTION]: Documentation in SKILL.md mentions the use of local Python scripts (scripts/cite.py and scripts/query.py) for querying citations. While these files are not included in the skill package, their described usage is limited to auxiliary CLI operations for the skill's primary function.
[PROMPT_INJECTION]: The instructions include robust constraints and 'HARD-GATE' sections intended to prevent model hallucinations and maintain doctrinal accuracy. These are functional constraints rather than attempts to bypass agent safety policies.
[EXTERNAL_DOWNLOADS]: Multiple references are made to fojin.app for accessing Buddhist source texts. These are static links for user or agent reference and do not involve the automated download or execution of external code.

master-xuyun