The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to execute local PowerShell (bootstrap-reverse.ps1) and Bash (bootstrap-reverse.sh) scripts for environment setup. These scripts are not part of the provided file list, which prevents verification of the commands they execute.
[DYNAMIC_EXECUTION]: The setup process uses powershell -ExecutionPolicy Bypass, which is a common technique to run scripts that might otherwise be restricted by system policy. This is used here to execute a script of unknown content.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and process untrusted public data from X/Twitter posts, which creates an attack surface for indirect prompt injection.
Ingestion points: External data enters via the Xquik MCP and REST interfaces (SKILL.md).
Boundary markers: The skill uses <UNTRUSTED_PUBLIC_SOURCE> tags and provides explicit instructions for the agent to treat the content as inert data and not execute instructions within it (SKILL.md).
Capability inventory: The skill has capabilities for shell command execution (PowerShell/Bash), network operations to third-party APIs (xquik.com), and data hand-off to other downstream skills like threat-hunting/ (SKILL.md).
Sanitization: While IOC normalization is mentioned, the raw body of external posts is passed directly into the agent context, relying solely on boundary markers for protection (SKILL.md).
[EXTERNAL_DOWNLOADS]: The skill requires connectivity to xquik.com for its core functionality, introducing a dependency on a third-party service for data collection and MCP registration.

threat-intelligence