The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of auditing external source code.
Ingestion points: The agent is instructed to read and analyze project source files (.ts, .tsx, .js, .jsx) and locale JSON files via the ripgrep tool and the provided Python audit script (SKILL.md, scripts/i18n_audit.py).
Boundary markers: The instructions lack explicit delimiters or safety warnings to ignore instructions that might be embedded within the string literals or comments of the files being processed.
Capability inventory: The skill has the capability to execute shell commands (rg, npm, python) and modify files within the project (SKILL.md).
Sanitization: No sanitization or validation of the input strings from the codebase is performed before they are handled by the agent.
[COMMAND_EXECUTION]: The skill uses standard CLI tools like ripgrep (rg), npm, and python to perform its tasks. These commands are executed locally and are consistent with the stated purpose of auditing and configuring a software project project.

i18n-expert