The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local Python script scripts/last30days.py. This script aggregates data from social platforms (Reddit and X) using available API keys to identify trending topics and engagement metrics.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources (Reddit, X, and web search results), which is a common surface for indirect prompt injection where external content attempts to hijack agent behavior.
Ingestion points: Data returned by the last30days.py script and the WebSearch tool.
Boundary markers: The instructions require the agent to synthesize findings and verify patterns across multiple sources before responding, which helps isolate research data from the agent's core instructions.
Capability inventory: The skill is authorized to use Bash (for data gathering), WebSearch, Read, and Write tools.
Sanitization: The skill employs strict output formatting templates and manual synthesis rules to ensure the agent focuses on facts and metrics rather than potentially malicious instructions embedded in the source material.

last30days