UltraCode — Claude Code Skill

Auto-activated for: TS/JS/Python/C#/Go/Rust/Java/C++/Swift/Kotlin/Zig/Bash code analysis

When to Use

ALWAYS use instead of Grep/Glob — 5-10x faster, understands meaning.

Task	Tool	Why
Search code by meaning	semantic_search	Understands natural language
Find similar code	find_similar_code	Semantic similarity
Impact analysis	analyze_code_impact	Shows what breaks
Find duplicates	find_duplicates / jscpd_detect_clones	Semantic + token detection
List entities in file	get_members	AST parsing
Show dependencies	list_entity_relationships	Dependency graph
Modify code safely	modify_code	With preview & rollback
Rename across project	rename_symbol	Updates all references
Diagnose stacktrace	analyze_stacktrace	Parse, resolve, suggest fixes
Architecture diagram	get_architecture_diagram	Mermaid/Graphviz/D2
DB schema	get_database_schema	SQL/ORM/Prisma/Redis
Security taint analysis	taint_analysis	SQL injection, XSS, missing auth
Detect anti-patterns	detect_patterns	100+ rules, 6 languages
Graph metrics	graph_metrics	PageRank, Louvain, centrality

Supported Languages

Language	Support	Features
TypeScript	Full	Type analysis, JSX/TSX
JavaScript	Full	ES6+, JSX, CommonJS/ESM
Python	Full	Type hints, async, decorators
Go	Full	Goroutines, interfaces
Rust	Full	Traits, lifetimes, macros
C#	Good	Roslyn, LINQ, async/await
Java	Good	Generics, annotations
C++	Good	Templates, namespaces
Swift	Good	Protocols, extensions
Kotlin	Good	Coroutines, data classes
Zig	Basic	Functions, structs, comptime
Bash	Basic	Functions, variables

---

Stacktrace Diagnosis

analyze_stacktrace

Parse and diagnose stacktraces from any language. Auto-detects language, resolves frames to code graph entities, classifies errors, runs backwards trace and impact analysis.

analyze_stacktrace({
  stacktrace: `TypeError: Cannot read property 'name' of undefined
    at processUser (src/users.ts:42:15)
    at handleRequest (src/server.ts:100:5)`,
  format: "text"         // text | json | mermaid
})

Supported languages: JS/TS (V8/Node), Python traceback, Java/Kotlin (Caused by chains), C#/.NET, Go goroutine panics, Rust panic/RUST_BACKTRACE, C/C++ (GDB/ASAN), Zig.

Param	Type	Default	Description
stacktrace	string	required	Full stacktrace text
language	enum	auto	Language hint
depth	number	10	Backwards trace depth
includeImpactAnalysis	boolean	true	Run impact analysis
includeBackwardsTrace	boolean	true	Run backwards trace
format	enum	"text"	text / json / mermaid
highlightRecentChanges	boolean	false	Annotate crash/call chain with recently-changed status + crash point history
recentCommitsCount	number	10	Recent commits to consider

Returns: error category, severity, crash location, call chain, callers, suggested fixes, Mermaid diagram. With highlightRecentChanges: recentChangeSummary + crashPointHistory.

---

Architecture Diagrams

get_architecture_diagram

Generate architecture diagrams in Mermaid, Graphviz DOT, or D2 format.

get_architecture_diagram({
  entryPoint: "src/server.ts",  // or omit for project overview
  depth: 2,                     // 1=files, 2=classes, 3=methods
  dataFlowLevel: 1,             // 0=none, 1=types, 2=params, 3=fields
  format: "mermaid"
})

---

Database Schema

get_database_schema

Show database schema from SQL, Prisma, ORM models (TypeORM, Sequelize, JPA, EF Core, Django, SQLAlchemy, GORM, Dapper, linq2db), and Redis key patterns.

get_database_schema({
  tableName: "users",          // partial match filter
  dbEngine: "postgres",       // postgres/mysql/clickhouse/redis/sqlite/mssql
  includeRelationships: true   // FK and code links
})

---

Security Tools

taint_analysis

Interprocedural taint analysis: trace untrusted data from sources to sinks.

taint_analysis({
  category: "missing_auth",  // sql_injection | xss | command_injection | path_traversal | ssrf | prototype_pollution | missing_auth
  offset: 0, limit: 10
})

---

Code Quality

detect_patterns

Detect anti-patterns, best-patterns, code smells, and optimization opportunities. 100+ rules across 6 languages. Supports highlightRecentChanges to annotate matches with recently-changed entity status.

detect_patterns({
  category: "anti-pattern",    // anti-pattern | best-pattern | code-smell | optimization
  tags: ["performance"],
  severity: "high",
  highlightRecentChanges: true  // annotate matches with Prolly Tree status
})

check_entity_patterns

Check specific entity for patterns with Big-O analysis.

analyze_state_chaos

Detect scattered state, race conditions, C# anti-patterns (async-void, mutable-static, god-service). Supports highlightRecentChanges to identify recently modified chaotic entities.

graph_metrics

Graph-based architecture metrics.

graph_metrics({
  metric: "pagerank",    // pagerank | louvain | centrality | bus_factor
  topN: 10,
  persist: true          // store in entity metadata
})

analyze_api_impact

Impact of API contract changes across Swagger/OpenAPI, Protobuf/gRPC, and GraphQL.

analyze_api_impact({
  contractType: "swagger",   // swagger | protobuf | graphql | auto
  changedFile: "api/openapi.yaml"
})

---

Code Modification via MCP

Automatic Validation

All code modifications (modify_code, add_member, create_file, rename_symbol) automatically run linting and error checking.

Safe Modification Workflow

1. semantic_search query="target code"       # Find what to modify
2. get_members filePath="file.ts"            # Get entity IDs
3. analyze_code_impact entityId="..."        # Check what breaks
4. create_snapshot description="Before..."   # Backup
5. modify_code entityId="..." preview=true   # Preview changes
6. modify_code entityId="..." preview=false  # Apply + auto-validate

Available Modification Tools

modify_code — Replace Entity Code

modify_code({
  entityId: "src/utils.ts::processData",
  newCode: `function processData(input: string): Result { ... }`,
  preview: true
})

rename_symbol — Smart Rename

rename_symbol({ entityName: "oldName", newName: "newName", preview: true })

add_member — Add to Class/Interface

add_member({ filePath: "file.ts", entityId: "MyClass", memberCode: "...", position: "end" })

create_file — Create New File

create_file({ filePath: "new-file.ts", content: "...", updateGraph: true })

Rollback & Safety

Tool	Purpose
create_snapshot	Backup before changes
undo	Rollback to snapshot
list_snapshots	View backups

---

History & Time Travel

Tool	Purpose
get_entity_history	Change history for entity across commits
diff_commits	Compare two graph commits
checkout_commit	View graph state at specific commit
list_commits	List graph version history

get_entity_history({ entityId: "src/auth.ts::login", limit: 10 })
diff_commits({ from: "abc123", to: "def456" })

---

Merge Tools

Tool	Purpose
semantic_merge	AI-powered 3-way merge
analyze_merge_conflicts	Predict conflicts before merge
get_merge_suggestions	AI suggestions for conflict resolution
get_semantic_merge_info	Merge capabilities info

---

Common Workflows

1. Understanding New Project

1. detect_technology_stack                   # Detect stack
2. get_graph_stats                           # View statistics
3. get_architecture_diagram depth=2          # Visual overview
4. analyze_hotspots metric="complexity"      # Find complex code
5. semantic_search query="entry point"       # Find entry points

2. Diagnosing an Error

1. analyze_stacktrace stacktrace="..." highlightRecentChanges=true  # Parse, diagnose, show recent changes
2. get_members filePath="crash_file.ts"     # Explore crash location
3. analyze_code_impact entityId="..." highlightRecentChanges=true   # What depends on crash point + recent changes

3. Security Audit

1. taint_analysis category="sql_injection"  # Find injection vulnerabilities
2. taint_analysis category="missing_auth"   # Find unprotected endpoints
3. detect_patterns category="anti-pattern" tags=["security"]

4. Finding Problematic Code

semantic_search minCyclomatic=10             # Complex code
semantic_search hasAwaits=true hasExceptions=false  # Async without try/catch
semantic_search hasDocumentation=false       # Undocumented

---

Tool Reference

Search

Tool	Purpose
semantic_search	Search by meaning (5-10x faster than Grep)
pattern_search	Multi-mode: entity/content/semantic/hybrid
find_similar_code	Find similar code by snippet
cross_language_search	Search across languages
find_related_concepts	Conceptually related code

Analysis

Tool	Purpose
analyze_code_impact	What breaks if I change X
analyze_hotspots	Complex code by complexity/changes/coupling
analyze_state_chaos	State management issues
analyze_api_impact	API contract change impact
analyze_stacktrace	Stacktrace diagnosis
detect_patterns	Anti-patterns, code smells, optimizations
graph_metrics	PageRank, Louvain, centrality, bus factor
taint_analysis	Security vulnerability detection
get_database_schema	Reconstructed DB schema

Entity & Graph

Tool	Purpose
get_members	List entities in file
list_entity_relationships	Entity dependencies
query	Natural language graph query
get_architecture_diagram	Architecture visualization

---

Related Skills

• ultracode-trace — Debugging, flow analysis, "why not called", stacktrace backwards trace
• ultracode-autodoc — Project memory with auto-updated code refs, find code by business meaning

---

Multi-Agent Worktree

Multiple AI agents can work in parallel on different branches using git worktrees.

Tool	Purpose
spawn_agent_worktree	Create worktree + prepare agent connection
list_worktree_agents	List active worktree sessions for current repo
cleanup_worktree	Remove worktree by branch name
get_worktree_info	Worktree/submodule/subtree detection info

// Create worktree for a new agent
spawn_agent_worktree({ branch: "feature/auth", baseBranch: "main" })

// Check active agents
list_worktree_agents({ includeInactive: true })

// Get full repo topology
get_worktree_info()

Key: All worktrees of the same repo share a single repoIdentity — no index duplication.

---

Cross-Project Support

index directory="D:\\other\\project"
semantic_search query="auth" projectPath="D:\\other\\project"

Storage: %LOCALAPPDATA%\UltraCode\projects\{hash}/