The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an Indirect Prompt Injection surface because it ingests and processes untrusted markdown content from various repository paths (skills, agents, teams, and guides) to perform translations.
Ingestion points: Source files read from skills/, agents/, teams/, and guides/ paths (Step 1.2).
Boundary markers: Absent. The procedure does not specify markers or instructions to differentiate between the data to be translated and potential instructions within the source files.
Capability inventory: The skill uses Write, Edit, and Bash tools, which provide the capability to modify the filesystem and execute local scripts (Step 2 and Step 6).
Sanitization: Absent. There is no requirement or step to sanitize or validate the input content for malicious instructions before processing.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute local scripts defined in the project's environment.
Evidence: Execution of npm run translate:scaffold in Step 2.1 and npm run validate:translations in the Validation section. These are vendor-provided scripts used for scaffolding and integrity checking.

translate-content