The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to recommend commands with elevated privileges (e.g., sudo chown -R $USER ~/.npm) and forceful process termination (e.g., kill -9 $(lsof -t -i:3000)) as standard fixes for permission and port conflicts.
[CREDENTIALS_UNSAFE]: For network and API errors (401/403), the instructions explicitly guide the agent to suggest checking .env files and verifying API keys, which targets sensitive local configuration files containing secrets.
[PROMPT_INJECTION]: The skill operates by ingesting and interpreting raw error messages, stack traces, and console logs provided by the user. Because this data often originates from external or third-party execution environments, it presents a surface for indirect prompt injection.
Ingestion points: User-supplied error messages and terminal output (SKILL.md).
Boundary markers: Absent; the skill does not define specific delimiters to isolate the error content from the agent's instructions.
Capability inventory: The skill is intended to output terminal commands and configuration changes for the user to execute.
Sanitization: Absent; there are no instructions to sanitize or validate the provided error text before interpretation.

error-fixer