Build semantic graphs from codebases using rpg-encoder and store results in .rpg/graph.json.
npx skills add https://github.com/userFRM/rpg-encoder --skill rpg
Please help me install this Agent Skill. Skill: rpg Source: https://github.com/userFRM/rpg-encoder/tree/main/.claude/skills/rpg Command: npx skills add https://github.com/userFRM/rpg-encoder --skill rpg
AI agents often struggle with raw codebases using grep/cat/find; RPG-Encoder provides a semantic, index-based view that makes code structure, behavior, and dependencies navigable in a single query.
Initialize the graph by running rpg-encoder build, then query results with rpg-encoder search.
High-intent search queries and answers about installing and using this skill.
You build a semantic graph by running the rpg-encoder build command, which uses Tree-sitter to parse the codebase. It stores the resulting dependency and entity structure in .rpg/graph.json for querying.
AI agents struggle with grep and cat because these tools lack structural context. A semantic graph index provides a navigable view of code behavior and dependencies, enabling accurate code path understanding in a single query.
Yes, you can trace dependencies and assess impact by running rpg-encoder search. It queries the pre-built semantic graph to fetch entity details and cross-file patterns, replacing manual multi-file searches.
Tree-sitter parsing extracts entity features across the repository to build a comprehensive semantic graph. This graph supports cross-file pattern detection and impact analysis by mapping dependencies navigable in a single query.
The best way to index code structure is building a semantic graph using Tree-sitter. This creates an intent-driven search index in .rpg/graph.json, allowing you to query architecture and dependencies instead of raw text.
After building the semantic graph with rpg-encoder build, use rpg-encoder search to query the stored graph. This allows you to fetch entity details, trace dependencies, and understand code paths by intent rather than text matching.
KakiseBuild a code relationship graph with tree-sitter and SQLite for semantic queries.
tedtv1007-ctrlAnalyze git diffs against a project knowledge graph to identify affected components and dependencies.
liujiayi3703Search codebases structurally using tree-sitter AST parsing.
massgenLocate text and structural code patterns across large codebases using ripgrep and ast-grep.
