Skip to content

Analysis

Tools for answering “what does this affect?” and “how did this evolve?“

Severity-scored blast radius. Run before any refactor.

code_impact(name="UserService", depth=3)

Returns CRITICAL / HIGH / MEDIUM / LOW dependents. See the full reference.


Most-called, most-complex, or largest symbols in the project.

code_hotspots(type="called") # most-called functions
code_hotspots(type="complex") # highest cyclomatic complexity
code_hotspots(type="large") # most lines of code

Good for finding where to focus optimization or test coverage.


Find code with no callers — potential dead code.

code_dead()

Returns symbols with zero inbound edges. Review before deleting — some are intentional entry points (CLI commands, event handlers).


Git diff plus all affected callers. Better than git diff.

code_diff_context()
code_diff_context(base="HEAD~3")

Shows changed lines AND every function in the graph that depends on what changed. Answers “what does this PR actually affect?“


Function-level git history. Better than git log.

code_changelog(name="processPayment")
code_changelog(name="UserService")

Shows commits that touched this specific function — not the whole file.


Full project overview: architecture, key modules, entry points.

code_summary()

Start here when exploring an unfamiliar project.


ASCII architecture diagram of the project.

code_map()

Shows modules and their relationships as a visual graph in text form.


Get the code around an error line number — useful when mapping stack traces back to source.

code_error_context(file="src/services/payment.ts", line=142)