The Agent Skills Directory

[SAFE]: The skill follows secure practices by instructing the agent to cite primary sources and use local, git-ignored storage for its research notes.
[SAFE]: The SKILL.md instructions focus on maintaining high standards for technical research, such as establishing evaluation criteria even when the user asks for a simple list, which promotes technical accuracy rather than security bypass.
[SAFE]: The evaluation tests (TESTS.md and eval.json) verify the skill's triggering logic and adherence to output formatting rules without introducing dangerous behaviors.
[SAFE]: References to external tools like 'Context7 MCP' are documented as standard extensions for fetching current documentation and do not involve unauthorized data exfiltration or remote code execution from untrusted sources.

research