The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to create symbolic links using ln -s on Linux/macOS and New-Item on Windows.
[COMMAND_EXECUTION]: The skill requests elevated privileges by explicitly instructing the agent or user to run PowerShell in Administrator mode on Windows.
[DATA_EXFILTRATION]: The skill's workflow involves scanning root directory configuration files, including package.json, pyproject.toml, and go.mod, to extract technology stack information.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of untrusted project metadata.
Ingestion points: The skill reads package.json, pyproject.toml, and go.mod from the project root.
Boundary markers: Absent; there are no delimiters or instructions to ignore instructions embedded within the scanned files.
Capability inventory: The skill possesses file system reading capabilities and command execution for creating symlinks.
Sanitization: Absent; the skill does not validate or sanitize extracted content before including it in the AGENTS.md documentation.

init