The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Bash command echo "${CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS:-not_set}" to detect the execution environment's capabilities (specifically if experimental agent teams are enabled). This is a benign environment check used to determine the appropriate workflow for spawning specialist agents.
[REMOTE_CODE_EXECUTION]: While the skill spawns multiple sub-agents using the Agent tool, this is a standard platform feature for complex tasks. The instructions provided to these agents are derived from the skill's own logic and user input, not from unverified remote sources.
[DATA_EXFILTRATION]: The skill uses WebSearch and WebFetch tools, but they are used legitimately to ground the systems analysis in real-world data about companies or policies. There is no evidence of sensitive data being sent to unauthorized third parties.

meadows