What problem does it solve? When working on the lci code indexer, it is hard to know which source files implement parsing and extraction for a given language, what lci actually extracts per language (symbols, references, imports, receiver types, side effects, catch sites), and why a file might produce zero symbols. This Skill provides a code map, capability matrix, and diagnostic recipes for those questions. ## Core Features & Use Cases - Code map of the parse pipeline: Locates the exact files and line numbers for language detection, the parser pool, UnifiedExtractor dispatch, and per-language reference/symbol handlers across 13 tree-sitter grammars plus Svelte script masking. - Per-language capability matrix: Shows which languages support symbols, references, cross-file imports, receiver-type call resolution, type relationships, side effects, and catch sites, including known gaps like Java/Kotlin/Ruby lacking cross-file import resolution. - Debugging zero-symbol files: Documents ParseSkipReason values, config knobs like max_parse_file_size, and probe recipes using lci status, symbols, refs, and targeted gtest filters. - Use Case: A developer adds a new grammar or fixes a Kotlin extraction bug and uses this Skill to find the fieldless-grammar workaround, the relevant dispatch sites, and the exact test filters to run. ## Quick Start Use the lci-parsing-languages skill to find where lci extracts symbols for Python and check why a file produced zero symbols.
