The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is instructed to automatically scan and ingest context from the entire project directory to build a naming brief.\n
Ingestion points: The skill uses Read, Glob, and Grep tools in the Project Context Scan section to ingest README.md, manifest files, marketing copy, and the entire docs/ directory.\n
Boundary markers: There are no instructions for the agent to use delimiters or ignore potentially malicious instructions embedded within these project files.\n
Capability inventory: The skill has access to shell command execution (via a Python script), WebSearch for market and domain analysis, and WebFetch for viewing registered domains.\n
Sanitization: No validation or sanitization of the ingested project data is performed before it is used to influence the agent's reasoning.\n- [COMMAND_EXECUTION]: The skill executes a local Python script to perform domain availability checks via the shell.\n
Evidence: SKILL.md contains the command: python "${CLAUDE_PLUGIN_ROOT}/skills/domain-hunter/scripts/domain_checker.py" name1 name2 --tlds .com,.io.\n
Argument Injection Risk: The skill is instructed to take user-supplied TLD lists and pass them directly to the script's --tlds flag without explicit sanitization, which could lead to command injection if the user provides malicious arguments.\n- [DYNAMIC_EXECUTION]: The skill dynamically assembles shell commands at runtime using generated names and user-provided inputs.\n
Evidence: The process for checking domains (Step 4) involves string interpolation of variable names and TLDs into a predefined Python execution command template.

brand-naming