The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the crates-io-pp-cli and crates-io-pp-mcp tools using go install from a specific GitHub repository and npx from a package registry. These resources originate from the vendor's own repository and package scope.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to install, verify, and run the CLI tool. This includes using the binary with various flags like --agent and output redirection sinks.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: Data is ingested from the external crates.io registry via the crates get and search subcommands in SKILL.md.
Boundary markers: The skill documentation does not specify the use of delimiters or boundary markers for the ingested ecosystem data.
Capability inventory: The skill allows writing output to arbitrary file paths using --deliver file:<path> and sending data to remote URLs via --deliver webhook:<url> as described in SKILL.md.
Sanitization: No specific sanitization or filtering of the ingested ecosystem data is mentioned in the instructions, presenting a potential surface for indirect injection from external crate metadata.

pp-crates-io