From darkroom
Provides token-efficient code analysis with call graphs, semantic search, impact analysis, dataflow graphs, and dead code detection. Auto-invokes for 'who calls X', 'what affects X', before refactoring or reading large files.
npx claudepluginhub darkroomengineering/cc-settingsThis skill is limited to using the following tools:
Token-efficient codebase analysis using llm-tldr v1.5+. **95% fewer tokens than reading raw files.**
Queries codebase knowledge graphs to search functions, trace callers/callees, list file entities, analyze impact, and run SurrealQL. Use for code structure, dependencies, and relationships.
Uses tree-sitter index for code navigation in Rust, Python, TypeScript, JavaScript, Go, Java, Scala, SQL: finds symbols, reads function implementations, traces callers, discovers tests.
Performs semantic search, grep, artifact fetching, relationship tracing, and data source listing across CodeAlive-indexed codebases and docs.
Share bugs, ideas, or general feedback.
Token-efficient codebase analysis using llm-tldr v1.5+. 95% fewer tokens than reading raw files.
Language is auto-detected (17 supported). No need to specify --lang unless overriding.
| Task | Command |
|---|---|
| "How does X work?" | semantic → context |
| "Who calls X?" | impact |
| "What would break?" | impact + change_impact |
| "Why is X null here?" | slice (backward) |
| "What does X affect?" | slice (forward) |
| "Project structure?" | arch + structure |
| "Find auth code" | semantic "authentication" |
| "Data flow in function" | dfg |
| "Control flow" | cfg |
| "Find dead code" | dead |
| "Type errors?" | diagnostics |
| "File tree" | tree |
| "Regex search" | search |
Find code by meaning, not exact text. Uses 5-layer embeddings (AST + call graph + CFG + DFG + PDG):
mcp__tldr__semantic { "project": ".", "query": "user authentication flow" }
mcp__tldr__semantic { "project": ".", "query": "error handling" }
Get LLM-ready summary instead of reading entire file:
mcp__tldr__context { "project": ".", "entry": "handleLogin", "depth": 2 }
Find all callers - critical before changing any function:
mcp__tldr__impact { "project": ".", "function": "useAuth" }
Understand project layers and dependencies:
mcp__tldr__arch { "project": "." }
What affects a specific line (backward) or what it affects (forward):
mcp__tldr__slice {
"file": "src/auth.ts",
"function": "login",
"line": 42,
"direction": "backward",
"variable": "user"
}
Cross-file function call relationships (language auto-detected):
mcp__tldr__calls { "project": "." }
Variable references and def-use chains:
mcp__tldr__dfg { "file": "src/auth.ts", "function": "validateToken" }
Basic blocks and branching:
mcp__tldr__cfg { "file": "src/auth.ts", "function": "handleRequest" }
Find tests affected by changed files (auto-detects from git diff):
mcp__tldr__change_impact { "project": "." }
Find unreachable code (language auto-detected):
mcp__tldr__dead { "project": "." }
Parse imports or find importers:
mcp__tldr__imports { "file": "src/utils.ts" }
mcp__tldr__importers { "project": ".", "module": "auth" }
Type checking and linting:
mcp__tldr__diagnostics { "path": "src/" }
Quick project structure overview:
mcp__tldr__tree { "project": "." }
mcp__tldr__tree { "project": "src/", "extensions": [".ts", ".tsx"] }
Search files by regex pattern:
mcp__tldr__search { "project": ".", "pattern": "TODO|FIXME|HACK" }
Functions, classes, methods per file (language auto-detected):
mcp__tldr__structure { "project": ".", "max_results": 50 }
Complete code structure from a single file (imports, functions, classes, call graph):
mcp__tldr__extract { "file": "src/auth.ts" }
Check uptime and cache statistics:
mcp__tldr__status { "project": "." }
1. tldr arch . # Get project overview
2. tldr semantic "feature name" # Find relevant code
3. tldr context functionName # Get LLM-ready summary
4. Read (only if more detail needed)
1. tldr impact functionName # Who calls this?
2. tldr change_impact # What tests affected?
3. tldr context functionName # Understand the function
4. Make changes
5. tldr diagnostics src/ # Check for errors
1. tldr slice file func line # What affects this line?
2. tldr dfg file func # Data flow analysis
3. tldr context func # Understand function
pipx install llm-tldr
tldr daemon start # Background service (300x faster)
tldr warm . # Build indexes including embeddings
tldr context BEFORE reading large filestldr impact BEFORE refactoringtldr semantic for "how does X work" questionsgrep ONLY for exact string matchinglanguage param — auto-detection handles 17 languagesReturn findings with: