The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell command execution via the dnx runner to perform all library inspection tasks, including querying types, members, and version differences.
[EXTERNAL_DOWNLOADS]: The skill uses dnx (described as functioning similarly to npx) to fetch and execute the dotnet-inspect package. This allows the agent to download and run code from external registries at runtime.
[INDIRECT_PROMPT_INJECTION]: The source --cat command enables the agent to fetch and display the contents of source files from remote SourceLink URLs. This introduces a surface where malicious instructions hidden in external source code could be ingested into the agent's context.
Ingestion points: Remote source code fetched from SourceLink URLs via the source command as documented in references/command-reference.md.
Boundary markers: The skill instructions do not specify any boundary markers or ignore-rules for the content retrieved from remote URLs.
Capability inventory: The skill possesses shell command execution capabilities through the dnx tool across all its functions.
Sanitization: There is no evidence of sanitization or filtering applied to the content fetched from external URLs before it is presented to the agent.

dotnet-inspect