svitovyd 0.1.5

Project map builder and structural query tool — index, find, trace, deps, idiff

svitovyd

Project map builder and structural query tool for codebases.

Scans any language codebase and produces a text map of:

• defined identifiers (classes, functions, endpoints, tables, …)
• cross-file references with relationship types (import / call / ref / expr)
• optionally variables and parameters (depth 3)

Query the map with find, trace (who calls?), deps (what depends on?), sym (health), idiff (structural diff).

Available as CLI and MCP server — works with Claude Code, opencode, nanocoder, 1bcoder, Cursor, and any MCP-compatible agent.

---

Installation

# CLI only
pip install svitovyd

# CLI + MCP (stdio)
pip install "svitovyd[mcp]"

# CLI + MCP + HTTP/SSE transport (for remote/LAN connections)
pip install "svitovyd[http]"

# CLI + web UI (Gradio)
pip install "svitovyd[ui]"

---

Quick start (CLI)

# Build map for current directory
svitovyd index .
svitovyd index . 3          # depth 3: also variables and parameters

# Search
svitovyd find auth
svitovyd find controller !test
svitovyd find "\UserService"          # blocks containing UserService
svitovyd find "\format_mismatch -format_mismatch"  # show only lines with term

# Trace call chain
svitovyd trace insertEmail            # who calls insertEmail?
svitovyd deps DatabaseManager         # what does DatabaseManager depend on?

# Health report
svitovyd sym --k 10

# Keyword vocabulary
svitovyd keywords index               # build .svitovyd/keyword.txt
svitovyd keywords extract "add author field to Book class" -f   # fuzzy extract
svitovyd keywords extract "rule" -l -o                          # like match + show files
svitovyd keywords                     # top-50 identifiers by reference count

# Structural diff (after changes)
cp .svitovyd/map.txt .svitovyd/map.prev.txt
svitovyd index .
svitovyd idiff --prev .svitovyd/map.prev.txt

---

Keywords

The keyword system bridges natural language task descriptions to real codebase identifiers. Two-step workflow:

Step 1 — build the vocabulary index:

svitovyd index .
svitovyd keywords index
# → .svitovyd/keyword.txt  (word, frequency, line numbers — CSV)

Step 2 — extract identifiers from a task description:

svitovyd keywords extract "add author field to Book class" -f
# → Book, author, addAuthor, BookRepository, …

Then use the extracted identifiers in find or trace:

svitovyd find "\Book"
svitovyd trace BookRepository

Extract flags

Flag	Meaning
-f / --fuzzy	Fuzzy subword match — splits camelCase/snake_case, matches if all query subwords (≥5 chars) appear in the identifier
-l / --like	Like match — any keyword containing the query token as a substring (%token%)
-n / --counts	Show frequency count next to each identifier: Book(47)
-s / --sort-count	Sort by codebase frequency descending (most common first)
-c / --csv	Comma-separated output (default: one per line)
-a / --alpha	Sort alphabetically (default: order of appearance in text)
-o / --origin	Show which files each keyword appears in: Book -> models/book.rb, spec/book_spec.rb

Exact match (default): query word must exactly match a keyword.txt entry — fastest, no false positives.

Like match (-l): "rule" matches parameter_rule, cop_rule, RuleIndex, load_rules — any identifier that contains rule as a substring. Stricter than fuzzy, broader than exact.

Fuzzy match (-f): "format mismatch" matches formatParameterMismatch, format_mismatch, FormatMismatch. Use for natural language task descriptions. Splits both query and keyword into subwords; query subwords shorter than 5 chars are treated as stopwords.

Origin grounding

-o (--origin) resolves each matched keyword back to the files it appears in — turning a list of identifiers into a list of files to investigate:

svitovyd keywords extract "add validation rule" -l -o
# parameter_rule -> lib/rubocop/cop/style/rule.rb, config/default.yml
# cop_rule       -> lib/rubocop/cop/base.rb, lib/rubocop/runner.rb
# load_rules     -> lib/rubocop/config_loader.rb

You can also look up a single keyword directly:

from svitovyd.query import keyword_to_files
files = keyword_to_files(".svitovyd/map.txt", "parameter_rule")
# ['lib/rubocop/cop/style/rule.rb', 'config/default.yml']

Ranked list (no task text)

svitovyd keywords           # top 50 by reference count
svitovyd keywords --k 100   # top 100
svitovyd keywords --plain   # one per line, for piping

---

Web UI

Browse the project map from a browser — useful when svitovyd runs on a remote machine or LAN server.

pip install "svitovyd[ui]"

svitovyd ui                        # opens at http://localhost:7860
svitovyd ui --port 7861            # custom port
svitovyd ui --map /path/to/map.txt # explicit map file

Binds to 0.0.0.0 by default, so it is immediately accessible from any machine on the same LAN:

http://192.168.1.42:7860

Tabs: Find · Trace · Deps · Sym · Keywords · Idiff · Download

The Keywords tab accepts an optional task description — enter text and it runs fuzzy identifier extraction; leave empty to get the ranked list. The Download tab lets you download map.txt and keywords.txt directly from the browser (useful when svitovyd runs on a remote server).

---

MCP server

The MCP server exposes 6 tools to any connected agent:

Tool	Description
map_index	Build or update the project map
map_find	Filter map blocks by filename/content
map_trace	BFS backwards — who calls this identifier?
map_deps	BFS forward — what does this depend on?
map_sym	Asymmetry and cohesion health report
map_idiff	Structural diff between two map snapshots

Start the server

Stdio (for local agents — Claude Code, 1bcoder):

svitovyd-mcp

HTTP/SSE (for remote agents — opencode, nanocoder, LAN):

svitovyd-mcp --http --port 8766
# or
svitovyd serve --http --port 8766

With a specific map file:

svitovyd-mcp --http --port 8766 --map-file /path/to/project/.svitovyd/map.txt

---

Setup per agent

Claude Code

Add to .claude/settings.json in your project or ~/.claude/settings.json globally:

{
  "mcpServers": {
    "svitovyd": {
      "command": "svitovyd-mcp",
      "args": []
    }
  }
}

With a specific map file:

{
  "mcpServers": {
    "svitovyd": {
      "command": "svitovyd-mcp",
      "args": ["--map-file", "/path/to/project/.svitovyd/map.txt"]
    }
  }
}

opencode

Start the HTTP server first:

svitovyd-mcp --http --port 8766

Add to your opencode config (~/.config/opencode/config.json or project-local):

{
  "mcp": {
    "svitovyd": {
      "type": "sse",
      "url": "http://localhost:8766/sse"
    }
  }
}

Or connect interactively if opencode supports it:

/mcp connect svitovyd http://localhost:8766/sse

nanocoder

Start the HTTP server:

svitovyd-mcp --http --port 8766

Add to nanocoder config:

{
  "mcpServers": {
    "svitovyd": {
      "url": "http://localhost:8766/sse"
    }
  }
}

1bcoder

/mcp connect svitovyd http://localhost:8766/sse

Or add to .1bcoder/mcp.yaml for auto-connect:

servers:
  - name: svitovyd
    url: http://localhost:8766/sse

Cursor

Add to .cursor/mcp.json in your project:

{
  "mcpServers": {
    "svitovyd": {
      "command": "svitovyd-mcp",
      "args": []
    }
  }
}

Any MCP-compatible client (generic)

Stdio transport — command: svitovyd-mcp

SSE transport — start server then connect to: http://localhost:8766/sse

---

MCP tools reference

map_index

path      directory to scan (default: ".")
depth     2 = definitions + links (default)
          3 = also variables and parameters
map_file  output path (default: <path>/.svitovyd/map.txt)

map_find

query     filter string — token syntax:
            term    filename contains term
            !term   exclude if filename contains term
            \term   include block if any child line contains term
            \!term  exclude block if any child line contains term
            -term   show ONLY child lines containing term
            -!term  hide child lines containing term
map_file  map file to query (default: .svitovyd/map.txt)

map_trace

identifier  function/class name defined in the codebase
depth       max BFS depth (default: 8)
map_file    map file to query

map_deps

identifier  function/class name OR file path substring
depth       max BFS depth (default: 8)
map_file    map file to query

map_sym

k           top-K hotspots for cohesion score (default: 5)
map_file    map file to query

map_idiff

prev_map_file   path to previous map snapshot (required)
map_file        current map file (default: .svitovyd/map.txt)

---

Supported languages

.py .js .ts .java .cs .go .rs .cpp .c .h .rb .php .kt .scala .swift .lua .html .css .jsx .tsx .vue .svelte .sql .yaml .yml .toml .json .xml .tf .hcl .sh .bat .ps1

---

Typical workflow with an AI agent

1. svitovyd index .                     # or: map_index path="."
2. svitovyd keywords index              # build keyword vocabulary
3. svitovyd keywords extract "task description" -f -c  # get real identifiers (fuzzy)
   svitovyd keywords extract "rule" -l -o              # like match + show origin files
4. map_find query="auth"                # explore by identifier / filename
5. map_trace identifier="login"         # understand call chain
6. map_deps identifier="DatabaseManager"# understand dependencies
7. map_sym                              # health check before refactoring
8. [make changes]
9. cp .svitovyd/map.txt .svitovyd/map.prev.txt
10. map_index path="."                  # rebuild
11. map_idiff prev_map_file=".svitovyd/map.prev.txt"  # verify impact

---

Part of the SIMARGL toolkit

svitovyd is one of five tools that together form an intellectual development support system:

Tool	Role
simargl	Task-to-code retrieval — given a task description, finds which files and modules are likely affected, using semantic similarity over git history
svitovyd	Project map — scans any codebase and produces a structural map of definitions and cross-file dependencies; exposes it as an MCP server
1bcoder	AI coding assistant for small local models — surgical context management, agents, parallel inference, proc scripts
yasna	Session memory — indexes conversations from all AI agents so you can find what was discussed, when, and where
radogast	Context drift monitor — measures how far an AI agent's conversation has drifted from the original task

• simargl answers: what code is related to this task?
• svitovyd answers: how is the code structured and what depends on what?
• 1bcoder answers: how do I work with local models efficiently?
• yasna answers: where did I already discuss this?
• radogast answers: is the AI agent still on track toward the goal?

Together they cover the full development loop: understand the codebase, find relevant history, work with AI locally, remember what was decided, and verify the context stays on target.

The name comes from Slavic mythology. Svitovyd (Світовид) is the four-faced god who sees all directions simultaneously — past, future, war, and harvest. A fitting name for a tool that maps an entire codebase at once.

---

About

(c) 2026 Stanislav Zholobetskyi, Oleh Andriichuk
Institute for Information Recording, National Academy of Sciences of Ukraine, Kyiv
PhD research: «Intelligent Technology for Software Development and Maintenance Support»