Completely reworked the explore approach:
- Claude (or keyword search) finds ONE entry point, not a list of symbols
- getCallGraph(entry, depth=3) traces the actual call chain deterministically
- No more AI-guessed symbol lists, bridge passes, or relevance filtering
- Search result clicks also trace the full call graph from that point
The graph data was always accurate — the problem was AI trying to guess
the whole flow. Now AI just finds the starting point, graph does the rest.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Stricter prompt rules: only symbols directly in the execution path,
every symbol must call or be called by the next, no tangential features.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Update Claude prompt to identify the entry point and return symbols in
execution order. The graph now centers on the entry point and auto-opens
its detail panel, giving users a clear starting point to trace the flow.
- Claude returns {entry, flow} instead of flat array
- Entry point is auto-selected and centered on load
- Detail panel opens immediately for the entry point
- Prompt asks for max 8-10 symbols in execution order
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Instead of expanding all callers/callees of seeds (which pulls in noise
from hub nodes like getSession), now:
1. Find direct edges between Claude's seeds
2. Only add non-seed nodes if they bridge 2+ isolated seeds
3. Cross-connection pass discovers hidden edges between result nodes
4. No more unrelated callers of hub nodes polluting the graph
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds `codegraph visualize` command that launches a localhost web UI for
visually exploring code relationships. Users can ask natural language
questions like "how does authentication work?" and see the relevant code
flow rendered as an interactive graph.
Key components:
- Visualizer HTTP server (src/visualizer/server.ts) with REST API
- Single-page frontend with Cytoscape.js graph + highlight.js code preview
- Claude CLI integration for intelligent query interpretation
- Dark theme, right-click context menus, keyboard shortcuts
- Detail panel with source code, callers, callees, hierarchy
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Traverses dependency graph to identify which test files depend on changed source files. Supports stdin input for git integration, custom test file patterns, and configurable traversal depth. Useful for targeted test execution in CI/CD pipelines.
Fixes#47 — "database is locked" after crash and MCP "not initialized"
when project IS initialized.
- FileLock: treat locks older than 10 minutes as stale regardless of PID
status, covering cases where PID was reused or kill signal check fails
- MCP server: log errors from tryInitializeDefault() to stderr instead of
silently swallowing, so transient open failures are diagnosable
- MCP server: retryInitIfNeeded() properly cleans up failed instances
before retrying, preventing resource leaks
- CLI: add 'codegraph unlock' command for manual lock file removal
Fixes#42 — tree-sitter can produce nodes with empty names (e.g. from
complex C/C++ declarators in header files). These nodes were silently
skipped at DB insert time, but their containment edges were still
inserted, causing a FOREIGN KEY constraint violation that crashed
indexing.
Two-layer fix:
- createNode() now returns null for empty names, preventing the node
and its edges from ever being created (Option A)
- storeExtractionResult() filters edges and unresolved refs to only
reference nodes that passed validation, as a safety net (Option B)
Fixes#54 — `codegraph init -i` crashes with "Fatal process out of memory: Zone"
on large codebases because all 16 tree-sitter WASM grammar modules were compiled
upfront by V8, exhausting the WASM Zone allocator.
Changes:
- initGrammars() now only initializes the tree-sitter WASM runtime (Parser.init()),
no longer eagerly loads all grammar files
- New loadGrammarsForLanguages() loads only grammars for languages actually present
in the project (e.g. a Dart project loads ~2-3 grammars instead of 16)
- Orchestrator detects needed languages after file scan, before parsing begins
- Embedding pipeline now uses quantized model (~67MB vs ~270MB) to further reduce
WASM memory pressure when embeddings are enabled
The serve command without --mcp prints a help banner. Writing this to
stdout breaks MCP stdio clients (like Cursor) that expect only JSON-RPC
on stdout. Move all banner output to stderr so stdout stays clean.
Fixes#43
command -v codegraph is unreliable inside npx because npx puts a
temporary binary in PATH. The check always passes, so the global
install is always skipped — which is the root cause of #37 and #38.
Fix: remove the check entirely, always run npm install -g.
- Revert configs (MCP, hooks) to use bare `codegraph` command
- Remove all npx references from configs and messaging
- Add preuninstall script that runs on `npm uninstall -g` to clean up
MCP server, permissions, hooks, and CLAUDE.md section
- Show uninstall instructions in post-install next steps
Global install is still attempted for bare `codegraph` convenience,
but now verifies the command is actually in PATH after install. If it
fails, users get clear actionable messages instead of silent swallowing.
Configs (MCP server, hooks) always use npx regardless — those never
break even if global install fails.
Remove the npm install -g attempt from the installer that silently
fails on many systems (permissions, PATH, node version managers).
All configs (MCP server, hooks, next-steps) now always use
npx @colbymchenry/codegraph. Global install offered as an optional tip.
Fixes#37, #38
- Fix defProc method lookup collisions when multiple classes share
method names (e.g. Create) by indexing qualified forms
- Remove overly broad 'Id' prefix filter that swallowed user-defined
symbols like Identifier or IdleTimer
- Use case-insensitive extension matching for DFM/FMX routing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Integrate main branch changes (WASM grammar architecture, centralized
resolution caches, SQLite adapter) with delphi-support branch. Pascal
grammar is now built as WASM and shipped in src/extraction/wasm/ for
consistency with the WASM-based grammar loading approach.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
web-tree-sitter has a known race condition when loading multiple WASM
grammars concurrently on Node.js 19+ (V8 10.8+). External scanner
symbols from one grammar can overwrite another's GOT entries, causing
"bad export type" errors for TypeScript, TSX, and other languages.
Replace Promise.allSettled(entries.map(...)) with a sequential for...of
loop so each grammar fully initializes before the next one starts.
Ref: https://github.com/tree-sitter/tree-sitter/issues/2338
Replace native tree-sitter with web-tree-sitter + tree-sitter-wasms for
universal cross-platform support. Add node-sqlite3-wasm as a fallback
when better-sqlite3 native bindings aren't available. Move better-sqlite3
and sqlite-vss to optionalDependencies so installs never fail.
Fix installer to use npx fallback when global npm install fails, so MCP
config, hooks, and quick-start instructions all work without the bare
codegraph command in PATH.
Fix tests: update schema version expectation, fix db test paths and
method names, extract MAX_OUTPUT_LENGTH as module constant, normalize
Windows path separators in import resolver.
Two major optimizations for the ref resolution phase:
1. Cache extractImportMappings() results per file path — previously
re-read and re-parsed the source file for every single ref from
that file (e.g. 100 refs from one file = 100 identical file reads)
2. Replace linear scan in matchFuzzy() with a lazily-built
case-insensitive Map index — O(1) lookup instead of iterating
all function/method/class nodes for every unresolved ref.
Also drop low-value prefix matching (confidence 0.3).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
These are project-specific directories and should be configured per
project in .codegraph/config.json, not hardcoded globally. Delphi-
specific excludes (__history, __recovery, *.dcu) remain as global
defaults since they are analogous to __pycache__ for Python.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace this.nodes.find() with a lazily-built Map<name, nodeId> for
matching implementation bodies to their declarations. Reduces per-file
complexity from O(methods × nodes) to O(nodes) for index build + O(1)
per lookup.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add _libs/**, libs/**, __history/**, __recovery/**, and *.dcu to
default exclude list. Delphi projects store external dependencies
in _libs/ or libs/ (often as Git submodules) and these should not
be indexed — same as node_modules for JS projects.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Two optimizations that eliminate the O(n²) bottleneck:
1. Add Pascal built-in filtering to isBuiltInOrExternal() — skips
resolution attempts for System.*, Vcl.*, Fmx.* and ~60 common
RTL identifiers (WriteLn, Create, Free, TObject, etc.)
2. Cache getNodesByKind() results during warm cache phase — matchFuzzy()
was calling the database 3 times per unresolved reference instead
of using the in-memory cache that warmCaches() already builds.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Some .dpr templates use "program;" without a name, which produces an
empty moduleName in the AST. Fall back to the filename (without extension)
to prevent nodes with empty names that cause downstream FK constraint errors.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The resolving refs phase stalled on large projects (3400+ files, 38k+ nodes)
because matchFuzzy loaded ALL functions/methods/classes per ref, import
mappings were re-extracted per ref, and fileExists hit disk every call.
Add kindCache, lowerNameCache, importMappingCache, and knownFiles set to
warmCaches(). Rewrite matchFuzzy to use O(1) lowercase index lookup instead
of 3x getNodesByKind scans. Cache import mappings per file. Pre-build file
existence set from the index for O(1) fileExists checks.
Fixes#28 - Python site-packages directories (e.g.
audio_tools/python/Lib/site-packages/) were not excluded by default,
causing massive index bloat and FOREIGN KEY failures when indexing
large libraries like tensorflow. The FK crash itself was already fixed
via INSERT OR IGNORE, but excluding these directories prevents the
bloat in the first place.
- Remove getNodesByIds, getNodesByKinds (zero callers)
- Remove getFileHashMap, getFileSyncMap (zero callers)
- Remove getDynamicStmt cache (only used by removed batch methods)
- Revert getStaleFiles to simple implementation (zero callers, no need to optimize)
- Remove intent field from SearchOptions (never read in search logic)
- extraction/index.ts: use picomatch with static import (replacing
dynamic require) and keep normalizePath for other call sites
- utils.ts: keep normalizePath from main, take PR's PID-based FileLock
Both functions have zero callers — dead code on arrival. Remove them
and their tests (9 tests) to keep the module focused on what's
actually used: search term extraction, path relevance scoring, and
kind bonuses.
- Remove extractFunctionVariable() and its dispatch (already handled by extractVariable)
- Remove dead getGrammar() export (zero callers)
- Deduplicate indexFile by delegating to indexFileWithContent
- Remove redundant arrow function variable extraction tests (covered by existing suite)
- Create file-kind nodes for each parsed source file
- Add isInsideClassLikeNode() for method vs function detection
- Extract arrow functions and function expressions from variable declarators
- Batch file I/O with FILE_IO_BATCH_SIZE=10 using Promise.all
- Add symlink cycle detection with visitedDirs Set in scanDirectory
- Add lazy grammar loading with exported getGrammar() function
- Add indexFileWithContent() for pre-read content processing
- Add tests for file nodes and arrow function extraction
- Add provenance column on edges for tracking how edges were created
- Add project_metadata table for version/provenance tracking
- Make unresolved_refs file_path/language NOT NULL with defaults
- Add composite indexes for unresolved_refs and edges.provenance
- Update v2 migration to handle all new schema additions
- Record schema version on initialize to prevent re-migration
- Add dynamic prepared statement cache (getDynamicStmt) for varying SQL
- Add batch methods: getNodesByIds, getNodesByKinds, getFileHashMap, getFileSyncMap
- Add project metadata methods: getMetadata, setMetadata, getAllMetadata
- Optimize getStats to use single aggregate query
- Optimize getStaleFiles to use temporary table JOIN
- Add provenance parameter to getOutgoingEdges
- Add intent field to SearchOptions type
- Add validateProjectPath() to reject sensitive system directories
- Add isPathWithinRoot/isPathWithinRootReal for symlink-aware path checks
- Replace hand-rolled glob-to-regex with picomatch to prevent ReDoS
- Add isSafeRegex() to reject custom patterns with nested quantifiers
- Replace FileLock with PID-tracking version that detects stale locks
- Add symlink detection in removeDirectory/listDirectoryContents
- Add subdirectory name validation in ensureSubdirectory
- Add atomicWriteFileSync and corrupted file backup in config-writer
- Add MCP input validation (validateString) for all tool handlers
- Fix CLAUDE.md section replacement to handle ### subsections correctly
Normalize paths to forward slashes in matchesGlob() and scanDirectory()
so glob exclude patterns work on Windows. Add getGitIgnoredDirectories()
using git ls-files to skip .gitignore'd directories during indexing.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Colby McHenryandClaude Opus 4.6
Mark HandGitHub
Olaf MonienandClaude Opus 4.6
Rafael Vescovi
Martin Oehlert