The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface
Ingestion points: The skill accepts a user-defined --query and processes external information gathered during the "Deep Research" phase (SKILL.md).
Boundary markers: The documentation does not specify the use of delimiters or specific instructions to the model to ignore embedded commands in the researched data.
Capability inventory: The skill executes a local Python script (scripts/research.py) which utilizes httpx for network communication and generates synthesized reports (SKILL.md).
Sanitization: No sanitization or filtering of external content is mentioned in the provided documentation.
[COMMAND_EXECUTION]: Local Script Execution
The skill instructions direct the agent to execute a local script python3 scripts/research.py with various arguments (SKILL.md). While this is the primary mechanism of the skill's functionality, it involves running executable code provided within the skill package.
[CREDENTIALS_UNSAFE]: API Key Management
The skill requires a GEMINI_API_KEY. The setup instructions correctly advise users to use environment variables or a .env file, which is considered a safe and standard practice for secret management (SKILL.md).

deep-research