The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill executes npx ai-devkit@latest lint, which downloads and runs the ai-devkit package from the NPM registry. This package corresponds to the primary tool suite described in the skill metadata.
[COMMAND_EXECUTION]: The skill uses several shell-based tools to gather context and analyze the repository, including npx, git status, git diff, and grep.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted data such as feature documentation and modified source code files, which presents a theoretical injection surface.
Ingestion points: External data is ingested when the agent reads feature docs and modified files (SKILL.md).
Boundary markers: No specific delimiters are defined to isolate the content of these files from the agent's instructions.
Capability inventory: The skill has access to shell execution via npx, git, and grep.
Sanitization: There is no evidence of content sanitization or validation for the files being reviewed.

dev-review