The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local terminal commands for project maintenance and validation tasks.
Evidence: Found in SKILL.md under the Validation section, which instructs the agent to run npm run lint and npm run test:unit.
Analysis: These are standard development commands used to verify code quality and functionality within the user's project. They are used appropriately for the skill's purpose and do not involve remote code or elevated privileges.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection through the processing of untrusted source code.
Ingestion points: The skill reads and edits .ts and .tsx files within the user's repository.
Boundary markers: Absent. There are no specific instructions to ignore natural language commands or instructions embedded within code comments or string literals in the files being processed.
Capability inventory: The agent has the ability to read and write files and execute terminal commands (npm run lint, npm run test:unit) in the project root.
Sanitization: Absent. The skill does not define validation or filtering for the code content it processes.
Analysis: This vulnerability surface is inherent to coding-assistant skills. While an attacker could place malicious instructions in source files, the skill's capabilities are restricted to standard development workflows, making the risk low.

paniolo-typescript-best-practices