The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/show-api.ps1 dynamically generates a temporary F# script file (probe.fsx) by interpolating parameters such as package name, version, and type. This file is then executed using the dotnet fsi command.
[REMOTE_CODE_EXECUTION]: The skill utilizes the F# interactive (#r) directive to download and load external libraries from the NuGet registry at runtime. This behavior is intended to allow the inspection of various package versions but involves downloading and executing code from an external repository.
[INDIRECT_PROMPT_INJECTION]: The skill processes package and member names which could potentially contain malicious payloads if the agent is instructed to inspect a crafted package name or version string.
Ingestion points: Input parameters to the scripts/show-api.ps1 script (found in SKILL.md and scripts/show-api.ps1).
Boundary markers: There are no specific boundary markers or instructions used to isolate untrusted data during processing.
Capability inventory: The skill can execute shell commands via dotnet fsi, create/delete files in the system's temporary directory, and perform network requests to restore NuGet packages.
Sanitization: The PowerShell script implements a ConvertTo-FSharpString helper to escape backslashes and double quotes before they are written into the generated .fsx file.
[COMMAND_EXECUTION]: The skill executes system commands including dotnet fsi, pwsh, and dotnet to perform its primary functions and run self-verification tests.

fsharp-verify-api