The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides setup instructions that involve downloading and executing remote scripts directly via a shell pipe, which is a high-risk pattern for executing unverified code.
Evidence: curl -fsSL https://d.officecli.ai/install.sh | bash for macOS/Linux systems.
Evidence: irm https://d.officecli.ai/install.ps1 | iex for Windows PowerShell environments.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external files (.xlsx, .csv), creating a potential attack surface where malicious content within these files could influence agent behavior.
Ingestion points: The skill frequently uses commands like officecli view, officecli get, and officecli query to extract data from user-provided workbooks.
Boundary markers: Absent. There are no instructions for the agent to treat ingested data as untrusted or to use specific delimiters to isolate it from instructions.
Capability inventory: The agent has extensive capabilities to modify the filesystem (create, set, add, remove, batch), read file contents, and generate HTML previews of spreadsheet data.
Sanitization: Absent. The skill does not define any validation or sanitization routines for the data parsed from external sources.
[DYNAMIC_EXECUTION]: The skill includes a template for a Python script (gen_batch.py) intended to generate JSON command batches at runtime for bulk data processing.
Evidence: The 'CSV / bulk import' section provides a Python snippet that dynamically constructs JSON objects for execution by the officecli batch command.

officecli-xlsx