The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation refers to a remote manifest accessed via the ts-pack list --manifest command to retrieve a list of supported languages.
[INDIRECT_PROMPT_INJECTION]: The skill describes capabilities for processing raw source text through the detect_language_from_content SDK function. This ingestion of untrusted external content represents a potential surface for indirect prompt injection if the agent attempts to follow instructions embedded in the analyzed code.
Ingestion points: SDK function detect_language_from_content and CLI parse/process commands.
Boundary markers: Not explicitly defined in the provided examples.
Capability inventory: CLI tools for parsing and structural extraction of code.
Sanitization: The skill does not describe specific sanitization steps for input code snippets.

detecting-languages