The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute multiple shell commands, including git clone, uv sync, and uv run strix. These commands are used to download, install, and operate the penetration testing environment.
[EXTERNAL_DOWNLOADS]: The skill clones the source code for the Strix tool from a remote repository at https://github.com/usestrix/strix.git. It further downloads dependencies via the uv sync command.
[INDIRECT_PROMPT_INJECTION]: As an autonomous penetration testing agent, this skill ingests data from external targets (web services, applications) and cloud API responses. This creates a vulnerability surface where a malicious application could return data containing instructions designed to manipulate the AI agent's behavior.
Ingestion points: Target application responses during active scanning, vulnerability report data from the Strix cloud API, and local configuration files such as ~/.strix/mcp-servers.json.
Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore potentially malicious content embedded within scan targets.
Capability inventory: The skill possesses the capability to execute shell commands, perform network operations against remote targets, and read/write local configuration files.
Sanitization: No evidence of sanitization, filtering, or validation of data received from external targets is present in the skill instructions.

strix