- --review at same commit hash skips to Phase 6, reuses existing graph
- Same commit hash without flags now asks user what they'd like to do
instead of silently stopping
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Update importCount metric description to reference batchImportData[file.path].length
instead of "number of import statements" (contradicted "do NOT extract imports" rule)
- Fix duplicate /index.js → /index.jsx in impl plan Task 5 extension variants
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Use CSS variables for diff edge colors in GraphView
- Make ReactFlow colorMode respond to theme preset
- Replace hardcoded text-white/bg-gray-900 in LearnPanel
- Use CSS variables for .kbd border and box-shadow
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Rename all gold-related CSS custom properties, Tailwind classes,
and keyframe animations from "gold" to "accent" to prepare for
the theme system where the accent color is user-configurable.
- --color-gold -> --color-accent (plus -dim and -bright variants)
- @keyframes goldPulse -> accentPulse
- .animate-gold-pulse -> .animate-accent-pulse
- All text-gold, bg-gold, border-gold, ring-gold Tailwind classes
No behavioral or visual change — pure mechanical rename.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a --messy mode to the large graph generator so robustness cases can be
tested with null fields, aliases, dangling refs, and other malformed data.
Update the dashboard to preserve non-fatal validation issues, log
auto-corrected and dropped items, and show a warning banner with expandable
details plus copy-to-clipboard support.
Ensures no alias value is itself an alias key, which would cause
single-pass normalization to produce a non-canonical value that
Zod then rejects. Catches this class of bug at test time.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The "tests" alias silently rewrites to "tested_by" without swapping
source/target, which inverts the relationship meaning and produces
incorrect edges. Direction-inverting aliases should fail validation
so the LLM gets explicit feedback to fix the edge.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
LLMs systematically abbreviate node types (e.g. "func" instead of
"function") and edge types (e.g. "extends" instead of "inherits"),
causing dashboard validation failures. This combines two fixes:
Option A: Rename the ambiguous `func:` ID prefix to `function:` across
all prompts, source code, tests, and example data so LLMs see consistent
naming. Also fix `relates_to` ghost edge type in django.md.
Option B: Add NODE_TYPE_ALIASES and EDGE_TYPE_ALIASES normalization maps
in schema.ts that transparently correct common abbreviations before Zod
validation, as a runtime safety net.
Closes#36
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Sync .claude-plugin/plugin.json (was stuck at 1.1.0) and .cursor-plugin/plugin.json (was 1.0.5) to match main version
- Update CLAUDE.md versioning section to list all four files that must be kept in sync
- Document scripts/generate-large-graph.mjs in CLAUDE.md
- Fix ? shortcut: add shiftKey: true so it matches on standard keyboards
- Fix Escape: keep hook always enabled so ESC closes the modal
(remove dead showKeyboardHelp branch, read all state at invocation time)
- Fix formatShortcutKey: don't show redundant ⇧ for shifted punctuation
- Remove stale showKeyboardHelp from useMemo deps
- Add "Press ? for keyboard shortcuts" hint in graph area
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Use dashboard store state at handler runtime and simplify shortcut wiring to avoid stale closures. App.tsx: replace many useDashboardStore selectors with runtime getState() calls inside keyboard shortcut actions and reduce useMemo dependencies to showKeyboardHelp. KeyboardShortcutsHelp.tsx: remove local Escape key effect (handled globally). useKeyboardShortcuts.ts: ignore shortcuts while typing in inputs/textareas/contentEditable (allow Escape), and improve Mac detection via userAgentData.platform fallback; use the same isMac flag in formatShortcutKey. README (zh-CN & ja-JP): add link to Turkish README (README.tr-TR.md).
Fixes#14 — dashboard becomes unusable with large knowledge graphs (2,700+ nodes).
- Web Worker for dagre layout: graphs above 200 nodes compute layout off the
main thread via a dedicated Web Worker, keeping the UI responsive
- Request ID correlation: concurrent layout calls are routed by incrementing
request ID, preventing race conditions where results cross-wire
- Worker error handling: onerror rejects pending promises; GraphView catches
errors and clears the loading spinner
- Cancellation cleanup: effect cleanup clears layouting state to prevent
stuck spinners when the graph switches below the async threshold
- Topology/visual split: dagre only re-runs when graph structure changes;
node selection, tour highlights, and search results are applied as a
cheap O(n) overlay without triggering relayout
- React.memo on CustomNode: prevents O(n) re-renders when selecting
individual nodes in large graphs
- Zero overhead for small graphs: below 200 nodes, the original synchronous
layout path is used unchanged
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Critical fixes:
- C1: Remove unused treeSitter.nodeTypes from configs and schema — extraction
logic is TS/JS-specific, nodeTypes gave false impression of generality
- C2+C3: Remove blocking temp dir prompt, default to .understand-anything/tmp/,
update all 5 prompt templates to use project-local paths instead of /tmp/
- C4: Fix false-positive framework detection — use "next": (with quotes) for
Next.js, remove cors/body-parser from Express keywords
Important fixes:
- I1: Change FrameworkConfig.language to languages[] array — React/Express/Vue
now correctly listed under both typescript and javascript
- I2: Fix getByExtension case sensitivity — .TS now resolves same as .ts
- I3: languageKeyFromPath returns null instead of throwing for unknown extensions
- I4: Duplicate framework registration is now a no-op instead of corrupting array
- I5: SKILL.md now says "skip silently" when language/framework snippet not found
Also: getForLanguage returns a defensive copy, 3 new tests added (157 total).
- Add blocking step in Phase 0 asking where to write temp files
($PROJECT_ROOT/.understand-anything/tmp/ or /tmp/), with explicit
STOP-and-wait language to prevent premature execution
- Make language/framework file discovery explicit in Phase 2 and
Phase 4 with concrete path examples and ls fallback, fixing
issue where Claude's Search tool couldn't find files in subdirs
- Clean up $TMP_DIR alongside intermediate files in Phase 7
Introduce keyboard shortcuts to the dashboard and add contributor docs. Adds a reusable useKeyboardShortcuts hook (with formatShortcutKey), a KeyboardShortcutsHelp modal component, and integrates shortcuts into App.tsx (registering keys like ?, Escape, /, ArrowLeft/Right, l, d and wiring store actions). Adds related styles (.glass-heavy, .kbd) to index.css. Also adds a comprehensive CONTRIBUTING.md with setup, workflow, testing, and PR guidelines.
Replace platform-specific install approaches with a consistent
per-skill symlink pattern for OpenCode and Codex (both share
~/.agents/skills/), matching the OpenCode docs requirement of
<name>/SKILL.md discovery.
Changes:
- OpenCode: migrate from opencode.json plugin config to direct
skill symlinks in ~/.agents/skills/<name>/
- Codex: replace broken bundle symlink (understand-anything/ had
no SKILL.md at root) with individual per-skill symlinks, matching
OpenCode exactly
- All platforms: add universal ~/.understand-anything-plugin symlink
so understand-dashboard SKILL.md can reliably find packages/dashboard/
- All platforms: use idempotent [ -e ] || [ -L ] || ln -s guard to
handle multi-platform installs on the same machine
- All platforms: use ln -sf and rm -f for robust re-installs and
partial uninstalls
- understand-dashboard/SKILL.md: replace fragile dirname-based path
resolution with a prioritized candidate loop with explicit error
guard
- Remove orphaned .opencode/plugins/understand-anything.js and its
build scaffolding
- Update all three README variants (en, zh-CN, ja-JP) to reflect
new OpenCode AI-driven install method
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace the TypeScript/JavaScript-biased analysis pipeline with a truly
language-agnostic system. The core architecture was already language-neutral
(graph schema, dashboard, search) — the bias lived in agent prompts,
tree-sitter plugin, and language-lesson system.
Core changes:
- LanguageConfig + FrameworkConfig types with Zod validation
- LanguageRegistry (12 languages) and FrameworkRegistry (10 frameworks)
- Config-driven TreeSitterPlugin replacing hardcoded TS/JS grammars
- PluginRegistry now delegates to LanguageRegistry for extension mapping
- Language-lesson system uses config for display names and concepts
Prompt system:
- SKILL.md generalized: dynamic injection of language snippets and
framework addendums instead of hardcoded if/else conditionals
- 12 language prompt snippets (languages/*.md) with concepts, patterns,
frameworks per language
- 10 framework addendums (frameworks/*.md) with canonical file roles,
edge patterns, architectural layers — Django/FastAPI/Flask preserved
and split, plus React/Next.js/Express/Vue/Spring/Rails/Gin added
- Extended entry points, directory patterns, and test patterns across
all 12 language ecosystems in base prompts
- Set minZoom to 0.01 so fit-view works on graphs with hundreds of nodes
- Add TourFitView to pan/zoom to highlighted nodes during guided tours
- Add SelectedNodeFitView to center graph on node selected via search
- Wrap GraphView in ReactFlowProvider to enable useReactFlow() hooks
Thanks to @blackboxaicode for the insight on these improvements.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The original commit 9fe003a introduced normalization logic in three
separate places (Phase 4 inline, Phase 5 inline, Phase 5.5), creating
an ordering conflict where "do not use field X" rules appeared before
"convert field X to field Y" rules. An LLM following instructions
sequentially could drop legacy fields before converting them.
- Consolidate all layer normalization into Phase 4 as ordered steps
- Consolidate all tour normalization into Phase 5 as ordered steps
- Remove redundant Phase 5.5 entirely
- Fix misleading "from current prompt" comments about envelope format
- Add defensive handling for nodes-as-objects edge case
- Wrap JSON examples in arrays for visual consistency with prompts
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>