skills/xberg-io/tree-sitter-language-pack/tree-sitter-language-pack-chunking-for-llms/Gen Agent Trust Hub
tree-sitter-language-pack-chunking-for-llms
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill describes the legitimate use of the 'tree_sitter_language_pack' library and 'ts-pack' utility for processing source code into syntactic chunks. The examples provided are standard for developer documentation and do not involve suspicious network, file, or command execution patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill documents a utility for processing external source code files, which constitutes an ingestion surface for potentially untrusted data.
- Ingestion points: Local source code files processed via the
ts-pack processcommand or thetree_sitter_language_pack.processPython function. - Boundary markers: The skill notes that chunks are returned as part of a structured JSON response under the
chunksfield. - Capability inventory: The skill utilizes a CLI tool (
ts-pack) and a Python library to read and tokenize source code files. - Sanitization: No specific content sanitization or escaping mechanisms are described in the provided documentation for the source code being chunked.
Audit Metadata