The Agent Skills Directory

[COMMAND_EXECUTION]: The profile configuration system (YAML) supports execution hooks such as on_start, on_complete, and on_error. These fields allow for the execution of arbitrary shell commands or scripts at various stages of the file transfer process.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to manage large-scale file transfers and includes configuration for LLM API keys (FASTCOPY_OPENAI_API_KEY, FASTCOPY_CLAUDE_API_KEY). This suggests the tool may process file content using AI, creating an attack surface where malicious content within transferred files could influence the agent's behavior. Mandatory evidence chain:
Ingestion points: Source file paths and directory structures (paths.source in YAML profiles).
Boundary markers: None identified in the provided instructions or configuration examples.
Capability inventory: The skill has extensive file system access, network capabilities (via curl in scripts), and shell command execution (via hooks).
Sanitization: No sanitization or validation of the content of transferred files is described.
[EXTERNAL_DOWNLOADS]: The CI/CD integration example (GitHub Actions) explicitly instructs users to download a binary from a placeholder URL (example.com), grant it execution permissions, and run it. While intended as a template, this encourages insecure practices regarding binary provenance and integrity verification.

fastcopy-file-transfer-tool