The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directly interpolates user-provided input ($ARGUMENTS) and user-influenced variables ($FILENAME) into bash commands: parallel-cli research run "$ARGUMENTS" and parallel-cli research poll "$RUN_ID" -o "$FILENAME". This pattern creates a risk of command or argument injection if a user provides a topic containing shell metacharacters (e.g., semicolons, pipes, or backticks).
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests and interprets data from external research results.
Ingestion points: The user topic ($ARGUMENTS) and the 'executive summary' output from the parallel-cli research poll command.
Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands within the research data.
Capability inventory: The skill uses the Bash tool to execute CLI operations and manage files.
Sanitization: No sanitization or validation of the research results is performed before the summary is shared with the user or used to inform subsequent agent actions.
[SAFE]: The skill references parallel-cli and parallel-web-tools. As these tools originate from the skill author, they are recognized as legitimate vendor resources and do not represent unauthorized third-party downloads.

parallel-deep-research