The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a local script audience-simulator.py with arguments derived from user input (e.g., {panel-id}, {slug}). This creates a potential command injection vector if the input is not sanitized before being passed to the shell.
[COMMAND_EXECUTION]: The skill accesses files and directories in the user's home path (~/.claude-marketing/). This involves reading brand profiles, guidelines, and agency SOPs, which could lead to unauthorized access to sensitive information if the skill logic is manipulated.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from both external files and user-provided message variants without explicitly defined boundary markers or sanitization steps.
Ingestion points: Brand profiles in ~/.claude-marketing/brands/, agency SOPs in ~/.claude-marketing/sops/, and user-provided message variants.
Boundary markers: Absent; the skill does not define delimiters or instructions to ignore embedded commands in the processed data.
Capability inventory: The agent can execute shell commands via the audience-simulator.py script.
Sanitization: No validation, escaping, or filtering of the external content or user-provided variants is specified before processing.

message-test