mcp-cpp-project-indexer is a deterministic C++ source-range indexer for
large, module-heavy projects and MCP-based AI code navigation.
It is not a compiler, LSP replacement, refactoring engine, semantic analyzer, or call-graph builder.
Its job is simple:
Find code. Read code. Do not guess code.
The indexer maps C++ symbols, files, and C++20 modules to exact source ranges so an AI can read only the code it needs.
Project and related AI orchestration work is documented on the MEF Programming homepage, including the ongoing relay/governance layer we are building around MCP tool use.
mcp-cpp-project-indexer builds a lightweight routing index over a C++ source
tree. MCP clients can then ask deterministic questions such as:
- where is this function/class/data member?
- which exact source range should be read?
- which module imports or exports this partition?
- which changed hunk intersects which indexed symbol or data range?
The indexer returns metadata and original source ranges. It does not claim to understand the program. The AI still has to read the returned source and reason from that evidence.
Minimal workflow:
User asks about Widget::OnScroll
-> find_symbol("Widget::OnScroll")
-> read_symbol(symbolId)
-> AI explains only what was visible in that source range
This keeps large C++ projects out of the prompt until exact source evidence is needed.
git clone https://github.com/walti1972/mcp-cpp-project-indexer.git
cd mcp-cpp-project-indexerpython <indexer-root>\build_project_index.py `
--root <project-root> `
--output-root <project-root>\.mcp-cpp-project-indexerThe generated index is written to:
<project-root>\.mcp-cpp-project-indexer
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexerFor multiple MCP clients or a long-running shared process, use HTTP transport:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 127.0.0.1 `
--http-port 8765Minimal LM Studio-style config:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"command": "python",
"args": [
"<indexer-root>\\code_index_mcp_server.py",
"--project-root",
"<project-root>",
"--index-root",
"<project-root>\\.mcp-cpp-project-indexer"
]
}
}
}Good first request:
Find the symbol Widget::OnScroll, read its implementation, and explain only
what is visible in the source range.
Expected tool path:
find_symbol -> read_symbol -> source-grounded answer
For best results, give your AI the rules from prompt_template.md. The short version is:
Use metadata to locate code. Read exact source ranges before explaining behavior.
Do not infer implementation behavior from metadata alone.
- 5-Minute Quick Start
- Production Scale & Performance
- TUI Control Center
- Why This Tool?
- Before / After
- How It Works
- Core Workflow
- What It Does
- Build And Update
- Project Discovery Config
- Control Center
- Start The MCP Server
- Client Configuration
- Possible Workflow Setups
- Command Line Reference
- Tool Overview
- Recommended AI Usage Rules
- Example Workflows
- Design Rules
- Development Backstory
- Smoke Tests
- Maintenance Checklist
This project is used on real C++ codebases, not only toy examples. Two recent scale runs show the intended range:
| Project | Files | Source lines | Lexer tokens | Symbols | Data declarations | C++20 modules | Full build |
|---|---|---|---|---|---|---|---|
| Anonymized commercial C++20 project | 7,046 | 979,658 | 4,682,882 | 97,924 | 36,551 | 3,754 | 19.5s |
| Chromium checkout | 137,622 | 30,792,607 | 137,365,399 | 2,327,255 | 818,188 | 0 | ~24m 32s |
These numbers are machine-dependent. The Chromium run used --jobs 60 on a
high-core workstation with an Intel Xeon Silver 4316 system, 128 GB RAM, and
enterprise NVMe SSD storage. It is a useful public stress test because it
exercises a very large classic include-based C++ codebase, while the anonymized
commercial project exercises dense C++20 module and partition metadata. The
Chromium run also validated the data/member indexer at scale: after fixing
nested-template >> depth handling, the public stress test surfaced 46,529
additional data declarations and 66,866 additional data-name aliases.
The SQLite-backed lookup index keeps server startup practical even at Chromium scale: the MCP server can start immediately and stay around 200 MB RAM after startup instead of loading millions of symbol/data/name entries into Python objects.
It is designed for workflows that combine the Codex desktop app or other MCP clients with Visual Studio navigation and, when needed, binary/decompiler evidence from tools such as IDA Pro.
In one measured workflow, exact source-range routing reduced source text read from roughly 2,000 lines to 283 lines, an 86% reduction.
For daily use, the indexer includes an optional mouse-capable TUI. It turns the project index into a small local control center:
- start the HTTP MCP server and watcher from one place
- run full builds, incremental updates, fast updates, and module-map rebuilds
- watch live server, watcher, lock, process, token, and index stats
- inspect build/update logs without switching tools
- toggle diagnostic file sections for deeper parser evidence when needed
Install the optional UI dependency and start the control center with explicit project/index paths:
pip install -r <indexer-root>\requirements-ui.txt
python <indexer-root>\indexer_tui.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--jobs 20 `
--http-url http://127.0.0.1:8765The UI is optional; the core indexer remains dependency-light and can still be driven entirely from scripts or MCP clients. For setup and keyboard shortcuts, see Control Center.
Large C++ projects are expensive to feed into an AI model when entire files are loaded just to find one function, class, import, or declaration. C++20 modules make this harder: many IDE/LSP-style tools still struggle with large module graphs, partitions, generated SDK headers, and build-specific configuration.
This indexer solves a narrower but very practical problem: it gives the AI a small, deterministic routing map. The AI can locate the relevant symbol, module, file, or changed hunk first, then read only the exact original source lines needed for the task.
Example from a real bug-finding workflow:
Whole file context: ~2000 source lines
On-demand source reads: ~283 source lines
--------------------------------------------
Reduction: ~86% less source text
| Standard AI code navigation | With mcp-cpp-project-indexer |
|---|---|
| ❌ AI reads whole files to find a symbol | ✅ AI asks for compact metadata, then reads the exact source range |
| ❌ Context fills with unrelated declarations and implementations | ✅ Context stays focused on the lines that matter |
| ❌ C++20 module consumers/imports are hard to route through | ✅ Module imports, re-exports, partitions, and consumers are exposed directly |
| ❌ Reviews start by scanning changed files manually | ✅ Change hunks are mapped to indexed symbol/data ranges |
| ❌ Tooling may imply semantic certainty it does not have | ✅ The indexer only returns routing facts and original source ranges |
The result is lower token usage, lower latency, less context drift, and more source-grounded analysis.
The indexer deliberately avoids pretending to be a compiler.
-
Fast token/structure scan Source files are scanned with a lightweight Python lexer and structural parser. The output is a deterministic table of contents: files, symbols, data declarations, lexical
#includedirectives, source ranges, diagnostics, and module facts. -
C++20 module map Module interfaces, partitions, imports, export-imports, and consumers are indexed so the AI can route through module-heavy code without asking an LSP to solve the whole build.
-
Incremental update and watcher The updater tracks content hashes and rewrites only changed index data where possible. The optional watcher can keep the MCP server cache fresh while you work in Visual Studio.
-
MCP tools with compact output controls Tools expose exact routing metadata first. The AI escalates only when needed: compact symbol lookup, file/module/change overview, exact
read_symbolorread_range, then deeper recursive source reads.
The indexer is only the table of contents. The AI performs recursive exploration and code review from the original source lines it explicitly reads.
Instead of this:
Read Renderer.cpp completely: ~2000 lines
use this:
find_symbol("Renderer::Paint")
read_symbol(symbolId)
inspect visible calls
read only relevant project callees
For changed-code review:
list_changed_files
get_file_change_hunks(includeIndexedRangeSummary:true, includeSource:false)
get_file_change_hunks(symbolId/dataId, includeSource:true)
read_symbol/read_range only when current source behavior is needed
The scanner extracts routing facts from C++ source files:
- files and stable file IDs
- C++20 modules and partitions
- lexical
#includedirectives - imports and exports
- namespaces
- classes / structs / enums
- functions / methods
- constructors / destructors / operators
- declarations and inline definitions
- exact
startLine/endLine - diagnostics for structurally suspicious files
It is stream/token based, not regex based.
Intentionally not included:
- no call graph
- no
find_references - no type resolution
- no template-instantiation resolution
- no compiler-accurate overload resolution
- no macro expansion
- no semantic summaries
- no bug analysis
- no
analyze_symbol(symbolId)
The AI should read source ranges and reason from the original code.
Default output directory:
<project-root>/.mcp-cpp-project-indexer/
Generated files:
.mcp-cpp-project-indexer/
manifest.json
files/
f_<pathHash>.json
index.sqlite
modules.json
diagnostics.json
update_state.json # written by build/update; used for fast incremental updates
module_map.json # generated by build_module_map.py
.watch_update_summary.json # temporary watcher/update summary
.update.lock # process lock for index writers
.watcher.lock # process lock for one active watcher
Global symbol and data routing indexes are stored in index.sqlite. The
per-file JSON indexes remain the source of truth for exact source ranges and
incremental rebuilds.
Optional JSONL export:
python <indexer-root>\export_index_jsonl.py --index-root <project-root>\.mcp-cpp-project-indexer --kind symbols --output symbols.jsonl
python <indexer-root>\export_index_jsonl.py --index-root <project-root>\.mcp-cpp-project-indexer --kind data --output data.jsonlScanner diagnostic file-index fields are emitted only with
--emit-diagnostics or --emit-diagnostic-file-indexes:
scopeIntervals
structuralEvents
functionBodyRanges
From any directory:
python <indexer-root>\build_file_index.py `
--file <project-root>\path\to\file.ixx `
--project-root <project-root> `
--output <project-root>\.mcp-cpp-project-indexer\diagnostic_file.jsonWith scanner diagnostic data:
python <indexer-root>\build_file_index.py `
--file <project-root>\path\to\file.ixx `
--project-root <project-root> `
--output <project-root>\.mcp-cpp-project-indexer\diagnostic_file.json `
--emit-diagnosticsIf --project-root is omitted, the file's parent directory is used.
Recommended usage from the C++ project root:
cd <project-root>
python <indexer-root>\build_project_index.pyThis writes to:
<project-root>/.mcp-cpp-project-indexer/
Explicit form:
python <indexer-root>\build_project_index.py `
--root <project-root> `
--output-root <project-root>\.mcp-cpp-project-indexerExample summary:
Built cpp.project_index.v1
Root: <project-root>
Output: <project-root>/.mcp-cpp-project-indexer
Files: 7076
Symbols: 97583
Names: 95674
Modules: 3774
Diagnostics: 7
Total code lines: 1750000
Total tokens: 14200000
SQLite index: <project-root>/.mcp-cpp-project-indexer/index.sqlite
Total tokens is the indexer's lexer token count over the indexed source after
comment blanking. It is a project-size metric, not an LLM billing-token count.
When the project root is inside a Git worktree and git is available, file
discovery respects Git ignore rules by filtering candidates through
git check-ignore --stdin. This excludes paths matched by .gitignore,
.git/info/exclude, or the user's global Git ignore file. Non-Git projects, or
systems without Git, fall back to the built-in excluded directory list.
Dot-directories such as .git, .vs, .cache, .idea, or .folder are
excluded by default.
For large projects with mixed source layouts, place indexer_config.json in
the project root or in any subdirectory. Config files are applied while walking
the tree: the root config becomes the base, and subdirectory configs can
override or extend it for that subtree.
Example:
{
"addExtensions": [".mm"],
"addExcludeDirs": ["generated", "third_party"],
"includeExtensionlessHeaders": true,
"useGitIgnore": false
}Supported fields:
{
"extensions": [".cpp", ".cc", ".h"],
"addExtensions": [".mm"],
"removeExtensions": [".c"],
"excludeDirs": ["out", "build"],
"addExcludeDirs": ["generated"],
"removeExcludeDirs": ["third_party"],
"includeExtensionlessHeaders": true,
"useGitIgnore": false
}extensions and excludeDirs replace the inherited values for that subtree.
add* and remove* fields modify the inherited values. Extensionless header
discovery is conservative and opt-in; it only accepts extensionless files whose
first lines look like C/C++ headers.
useGitIgnore:false disables the final git check-ignore --stdin pass for
very large repositories where Git ignore filtering is more expensive than an
explicit indexer config.
After a full index build, changed files can be detected and re-indexed incrementally.
Dry-run:
python <indexer-root>\update_project_index.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--dry-runFast update for already-indexed files:
python <indexer-root>\update_project_index.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--known-files-onlyWatcher-style update for a known changed file:
python <indexer-root>\update_project_index.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--known-files-only `
--changed-file path\to\changed.cpp--known-files-only skips full discovery of new files. This is ideal for save/watch loops.
--changed-file can be repeated and lets a watcher avoid hashing unchanged files.
Index writes are protected by an exclusive .update.lock file in the index
root. This prevents full builds, incremental updates, and module-map rebuilds
from writing the same index files at the same time.
The standalone watcher polls source files, debounces changes, runs the incremental updater,
and optionally rebuilds module_map.json.
python <indexer-root>\watch_project_index.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--jobs 20For pure file modifications, the watcher calls:
update_project_index.py --known-files-only --changed-file <path>
This hashes only the changed candidate file. If the content hash did not change, the watcher skips module-map rebuild and MCP cache reload work.
Only one watcher should own an index root. The watcher uses .watcher.lock and
exits if another watcher is already active for the same index.
Diagnostics are non-fatal. They indicate best-effort structural warnings for individual files.
Print diagnostics summary:
python -c "import json; from collections import Counter; d=json.load(open(r'<project-root>\.mcp-cpp-project-indexer\diagnostics.json',encoding='utf-8')); print(len(d)); print(Counter(x.get('code') for x in d)); [print(x['relativePath'], x['code'], x['message'], x.get('range')) for x in d]"After building the project index:
python <indexer-root>\build_module_map.py `
--index-root <project-root>\.mcp-cpp-project-indexerOutput:
<project-root>/.mcp-cpp-project-indexer/module_map.json
The module map contains:
- module names
- primary modules and partitions
- files defining each module
- imports
- importedBy
- a module tree
- unresolved imports
It is metadata only. It does not analyze implementation behavior.
The core indexer has no third-party runtime dependencies. For day-to-day operation there are two optional terminal control surfaces:
indexer_tui.py: a polished mouse-capable Textual UIindexer_control.py: a dependency-free command-line fallback
Install the optional UI dependency:
pip install -r <indexer-root>\requirements-ui.txtStart the Textual UI:
python <indexer-root>\indexer_tui.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--jobs 20 `
--http-url http://127.0.0.1:8765The Textual UI provides a real terminal interface with buttons, mouse support,
status cards, live HTTP /status polling, activity logs, and process control
for build/update/watch/server actions.
Mouse input depends on the terminal emulator. On Windows, use Windows Terminal or another modern terminal that forwards mouse events to terminal applications. Keyboard shortcuts work everywhere Textual can run.
UI settings are saved under <indexer-root>/.ui-settings/ on exit and when
diagnostic file sections are toggled. Each settings file is keyed by project
name plus a path hash, so UI preferences do not write into the project root or
generated index directory. Stored preferences include the HTTP URL, jobs value,
diagnostic-section toggle, management API launch settings, active Textual
theme, and project/index paths.
The settings file can also be edited directly for external Relay UI setups:
{
"httpUrl": "http://127.0.0.1:8765",
"jobs": 20,
"managementApiEnabled": true,
"managementToken": "local-secret-token",
"emitDiagnosticFileIndexes": true,
"theme": "textual-dark"
}The Textual UI has a separate Start HTTP + management action. It starts the
HTTP MCP server with watcher and --enable-management-api, using the saved
managementToken when configured. This action is intentionally local to the
TUI and is not exposed as a management command on the HTTP API.
Fallback control center:
python <indexer-root>\indexer_control.py `
--root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--jobs 20 `
--http-url http://127.0.0.1:8765Common keys:
B full build
U incremental update
F fast known-files update
M rebuild module map
H start HTTP server with watcher
G start HTTP server with watcher and management API
W start standalone watcher
S toggle diagnostic file sections for launched commands
X stop the currently running command
R refresh
Q quit
When the HTTP server is running, both control centers poll /status for live
server, watcher, lock, and index stats. If the HTTP server is not reachable,
they fall back to reading manifest.json and module_map.json from disk.
Text tree:
python <indexer-root>\dump_module_tree.py `
--index-root <project-root>\.mcp-cpp-project-indexerWrite to file:
python <indexer-root>\dump_module_tree.py `
--index-root <project-root>\.mcp-cpp-project-indexer `
--output <project-root>\.mcp-cpp-project-indexer\module-tree.txtImport tree for one module:
python <indexer-root>\dump_module_tree.py `
--index-root <project-root>\.mcp-cpp-project-indexer `
--imports Example.Module:Partition `
--max-depth 5The server reads an existing index. It does not rebuild the index.
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexerThe server is read-only and exposes locator/read tools over MCP stdio.
To let the server keep its in-memory cache current during editing, start it with the built-in watcher:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--watch-index `
--watch-jobs 20The server watcher writes all progress to stderr so stdout remains valid MCP JSON-RPC. After a real index change, it reloads the server cache automatically. If a save only changes mtime and the content hash is unchanged, it skips module map rebuild and cache reload.
If another watcher already owns the same index root, the MCP server continues read-only and does not start its own watcher. This avoids concurrent writer processes when multiple MCP clients start separate stdio server instances.
For setups where multiple MCP clients would otherwise start separate stdio server processes, the server can also expose the same JSON-RPC MCP surface over HTTP:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 127.0.0.1 `
--http-port 8765 `
--watch-index `
--watch-jobs 20The HTTP endpoint is:
POST http://127.0.0.1:8765/mcp
Status endpoints:
GET http://127.0.0.1:8765/health
GET http://127.0.0.1:8765/status
This keeps one long-running server process, one in-memory index cache, and one watcher for the index root. Clients that only support stdio still need a stdio configuration or a small client-side bridge to this HTTP endpoint.
External control UIs can optionally enable a small management surface on the same HTTP server. It is disabled by default because it can start build/update processes.
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 127.0.0.1 `
--http-port 8765 `
--enable-management-api `
--management-token <token>Endpoints:
GET /management/status
GET /server/management/capabilities
POST /management/command
GET /management/log?since=<eventId>&limit=<n>
GET /management/log/stream
GET /management/server-log?since=<eventId>&limit=<n>
GET /management/server-log/stream
/server/management/capabilities returns the machine-readable
intent/category/tool contract used by external relay/governance layers. The
legacy alias /management/capabilities is also accepted.
/management/status includes a dashboard object shaped for external control
centers. It contains the same high-value fields shown by the TUI:
dashboard.project.text
dashboard.index.text
dashboard.server.url / pid / ramText / cpuTimeText / cpuText / uptimeText
dashboard.watcher.runningText / lockText / last
dashboard.counts.filesText / symbolsText / dataText / modulesText / diagnosticsText
dashboard.stats.codeLinesText / tokensText / cpuTimeText / threadsText
dashboard.locks.updateFile / watcherFile
dashboard.mode.diagnosticFileSectionsText / jobsText / theme
Raw numeric values are included next to the formatted *Text fields where the
server can provide them. dashboard.mode.theme is reserved for external UIs;
the HTTP server itself does not own a visual theme.
The generic server.process and management.runner.process objects also expose
normalized CPU fields when available:
cpuUserSeconds
cpuSystemSeconds
cpuTimeSeconds
cpuTimeText
cpuCoresAverage
cpuPercentMachine
cpuText
When configured, the token protects the shared HTTP MCP/status surface as well as management endpoints. Pass it as either:
Authorization: Bearer <token>
x-api-key: <token>
Commands are single JSON objects:
{ "command": "build", "jobs": 20 }
{ "command": "update", "jobs": 20 }
{ "command": "fast_update", "jobs": 20 }
{ "command": "module_map" }
{ "command": "reload_index" }
{ "command": "start_watcher", "jobs": 20 }
{ "command": "stop_watcher" }
{ "command": "stop_command", "wait": true }/management/log returns management command output. /management/log/stream
is the matching SSE stream. Build and update subprocess output is read
asynchronously, so a UI can keep polling status and streaming logs while large
projects are being indexed.
/management/server-log returns recent HTTP/MCP traffic log events, including
request/response byte counts and the MCP method detail where available.
/management/server-log/stream is the matching SSE stream for live request
activity. MCP tools/call events also include a compact structured mcp
object with toolName, argumentKeys, and bounded arguments for expandable
UI details. After the JSON-RPC response is known, the event includes
mcp.outcome (success or error) and mcp.errorCount so control UIs can
color failed tool calls without parsing the response payload.
HTTP keep-alive requests reset request-local log metadata, so status polling
cannot inherit a previous MCP call label.
Local development can keep using loopback HTTP:
--transport http --http-host 127.0.0.1 --http-port 8765
For customer or LAN deployments, do not expose the server as plain HTTP. The server refuses to start on a non-loopback bind address unless TLS and authentication are enabled.
Security profiles:
| Profile | Intended use | Requirements |
|---|---|---|
local-dev |
same-machine development | loopback HTTP allowed |
trusted-lan |
Relay and indexer on different trusted hosts | TLS plus token or mTLS |
production |
customer deployment | TLS plus token or mTLS |
Token-authenticated TLS example:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 10.0.0.20 `
--http-port 8765 `
--management-security-profile trusted-lan `
--management-tls cert `
--management-cert security\indexer-server.crt `
--management-key security\indexer-server.key `
--management-token <token> `
--management-cors-origin https://relay.customer.internal `
--management-allow-ip 10.0.0.12mTLS example:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 10.0.0.20 `
--http-port 8765 `
--management-security-profile production `
--management-tls cert `
--management-cert security\indexer-server.crt `
--management-key security\indexer-server.key `
--management-client-ca security\relay-clients-ca.crt `
--management-require-client-cert `
--management-cors-origin https://relay.customer.internal `
--management-allow-ip 10.0.0.12For encrypted local testing without a customer certificate, a self-signed certificate can be generated on first start:
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexer `
--transport http `
--http-host 127.0.0.1 `
--http-port 8765 `
--management-tls self-signed `
--management-auto-certThe generated files are stored below:
<index-root>\certs\management-cert.pem
<index-root>\certs\management-key.pem
Self-signed TLS encrypts traffic but is not automatically trusted by browsers or remote clients. Customer deployments should use a customer CA, public certificate, or explicit mTLS trust material.
Example mcp.json:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"command": "python",
"args": [
"<indexer-root>\\code_index_mcp_server.py",
"--project-root",
"<project-root>",
"--index-root",
"<project-root>\\.mcp-cpp-project-indexer"
]
}
}
}For a shared HTTP server without token authentication:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"url": "http://127.0.0.1:8765/mcp"
}
}
}If the HTTP server was started with --management-token <token>, LM Studio must
send that token as a header:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"url": "http://127.0.0.1:8765/mcp",
"headers": {
"Authorization": "Bearer <token>"
}
}
}
}The server also accepts X-API-Key:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"url": "http://127.0.0.1:8765/mcp",
"headers": {
"X-API-Key": "<token>"
}
}
}
}After rebuilding the index or module map, restart the MCP server / LM Studio.
For Codex, configure this repository as an MCP stdio server and put the agent rules where Codex will load them as project instructions.
Example MCP server entry:
[mcp_servers.mcp-cpp-project-indexer]
command = "python"
args = [
"<indexer-root>\\code_index_mcp_server.py",
"--project-root",
"<project-root>",
"--index-root",
"<project-root>\\.mcp-cpp-project-indexer",
]The prompt rules from prompt_template.md should be copied into
an AGENTS.md file in the project root that Codex opens, for example:
<project-root>\AGENTS.md
Codex reads AGENTS.md as repository/project instructions for the current working
tree. If you usually start Codex from a parent workspace instead of the C++ project
root, put the file in that workspace root or open Codex directly in the C++ project
root so the instructions are in scope.
Keep the MCP server configuration and the prompt rules separate:
- MCP config starts the tool server and points it at the generated index.
AGENTS.mdtells the agent how to use the tools safely and compactly.
After changing the MCP server configuration or rebuilding the index, restart the MCP server / Codex session so the updated tool schema is visible.
Claude Code and Claude Desktop use MCP slightly differently, so keep the setup split by client.
For Claude Code, put the MCP server configuration in a project-scoped .mcp.json
file at the project root if the setup should travel with the repository:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"type": "stdio",
"command": "python",
"args": [
"<indexer-root>\\code_index_mcp_server.py",
"--project-root",
"<project-root>",
"--index-root",
"<project-root>\\.mcp-cpp-project-indexer"
],
"env": {}
}
}
}Equivalent CLI setup:
claude mcp add mcp-cpp-project-indexer --scope project -- `
python <indexer-root>\code_index_mcp_server.py `
--project-root <project-root> `
--index-root <project-root>\.mcp-cpp-project-indexerCopy the rules from prompt_template.md into Claude Code's project memory file:
<project-root>\CLAUDE.md
Claude Code also supports .claude/CLAUDE.md; use that path if you prefer to keep
Claude-specific files under .claude/. The important part is that the file is in
the project root that Claude Code opens, otherwise the tool-use rules may not be
loaded.
Claude Desktop uses its application MCP configuration file. On Windows this is usually:
%APPDATA%\Claude\claude_desktop_config.json
Add the server under mcpServers:
{
"mcpServers": {
"mcp-cpp-project-indexer": {
"type": "stdio",
"command": "python",
"args": [
"<indexer-root>\\code_index_mcp_server.py",
"--project-root",
"<project-root>",
"--index-root",
"<project-root>\\.mcp-cpp-project-indexer"
],
"env": {}
}
}
}Claude Desktop does not automatically read repository instruction files like
CLAUDE.md for arbitrary chats. Paste the relevant rules from
prompt_template.md into the project/chat instructions you use
with Claude Desktop, or use Claude Code when you need repository-scoped persistent
instructions.
After changing .mcp.json, claude_desktop_config.json, or rebuilding the index,
restart the Claude client so the updated tool schema is visible.
A practical review workflow is to use mcp-cpp-project-indexer as the primary
navigation layer and let the AI reason only from exact source ranges it has read.
Typical flow:
1. User asks:
Review module X, file X, or function X for bugs.
2. AI uses mcp-cpp-project-indexer:
find module/file/symbol
read exact source ranges
follow only relevant project-local calls
avoid whole-file reads unless needed
3. AI reports findings:
file path
line range
source-grounded explanation
uncertainty where more context is needed
4. AI uses Visual Studio MCP only after analysis:
open the file
navigate to the exact finding location
This keeps the model context clean. The indexer handles routing and token reduction; the AI performs the analysis from original source lines; Visual Studio is used as the developer-facing editor handoff.
In C++20 module-heavy projects, Visual Studio/IntelliSense/clangd-style tooling may fail to resolve module symbols reliably enough for AI navigation.
Recommended role split:
mcp-cpp-project-indexer
Primary navigation layer.
Use for symbols, files, modules, source ranges, imports, imported-by metadata.
Visual Studio MCP
Editor and project-state layer.
Use for opening files, jumping to locations, looking at build/output/editor state.
Do not use as the primary C++20 module symbol resolver.
IDAPro MCP
Binary/decompiler layer.
Use only when source evidence is insufficient, or for ABI, crash, reverse
engineering, decompiled code, imports/exports, or binary-behavior questions.
Prompt rule:
Use mcp-cpp-project-indexer first for C++ source navigation.
Use Visual Studio MCP only when IDE/editor/build state is needed.
Use IDAPro MCP only when the question requires binary or decompiler evidence.
For code that interacts with undocumented Windows components, source evidence may not be enough. A useful setup is to keep the relevant DLL loaded in IDAPro and let the AI inspect decompiled/disassembled code only when source-level evidence leaves behavior unclear.
Example scenario:
Project code uses wrappers around undocumented DirectUI behavior.
IDAPro has dui70.dll loaded.
1. AI uses mcp-cpp-project-indexer first:
find the project wrapper/function
read the exact source range
follow only relevant local calls
2. If behavior is still unclear:
use IDAPro MCP to inspect the corresponding function, import, export,
vtable target, or decompiled implementation in dui70.dll
3. AI combines evidence:
source callsite and wrapper behavior
binary/decompiler evidence from the undocumented implementation
final finding with source line ranges and binary evidence notes
4. AI uses Visual Studio MCP only for handoff:
open the project source file
navigate to the line that should be reviewed or changed
This keeps reverse-engineering work targeted. The AI does not browse the binary blindly; it enters IDAPro with a source-grounded question. That reduces irrelevant context, lowers token usage, and helps the model keep the source/binary reasoning chain intact.
The expensive part of AI code review is often not reasoning; it is locating the small amount of source that actually matters. The indexer reduces the navigation problem to compact metadata and exact source ranges. Other tools can then stay focused on what they are good at: IDE interaction, build state, or binary facts.
Build one per-file index JSON.
--file PATH C++ source/module file to index. Required.
--project-root PATH Root used for normalized relative paths.
--output PATH Output JSON path.
--output-root PATH Directory used when --output is omitted.
--case-insensitive-paths / --no-case-insensitive-paths
Case-fold relative paths before hashing. Default: true.
--blank-comments / --no-blank-comments
Blank comments before scanning while preserving lines.
--emit-diagnostics / --no-emit-diagnostics
Include scanner diagnostic data.
--print-summary-json Print summary JSON.
Build the full project index.
--root PATH Project/source root. Default: current directory.
--output-root PATH Index output root. Default: <root>/.mcp-cpp-project-indexer.
--extensions EXT [EXT ...] Source extensions, e.g. .cpp .cc .mm .h .ixx or cpp,h,ixx.
--include-extensionless-headers / --no-include-extensionless-headers
Also discover extensionless files that look like
C/C++ headers. Uses a conservative first-lines
heuristic. Default: false.
--git-ignore / --no-git-ignore Filter discovered files through git check-ignore
when available. Default: true.
--exclude-dir NAME Extra excluded directory name. Repeatable or comma-separated.
--case-insensitive-paths / --no-case-insensitive-paths
Case-fold relative paths before hashing. Default: true.
--blank-comments / --no-blank-comments
Blank comments before scanning. Default: true.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Include scanner diagnostic fields in files/<fileId>.json.
Compatibility alias: --emit-debug-file-indexes.
Default: false.
--print-summary-json Print summary JSON.
--list-defaults Print default extensions and excluded directories.
--jobs N Worker processes. 1 = sequential, 0 = conservative auto.
--progress / --no-progress Show progress on stderr. Default: true.
Incrementally update an existing project index.
--root PATH Project/source root. Default: current directory.
--index-root PATH Index root. Default: <root>/.mcp-cpp-project-indexer.
--extensions EXT [EXT ...] Source extensions for discovery.
--include-extensionless-headers / --no-include-extensionless-headers
Also discover extensionless files that look like
C/C++ headers. Uses a conservative first-lines
heuristic. Default: false.
--git-ignore / --no-git-ignore Filter discovered files through git check-ignore
when available. Default: true.
--exclude-dir NAME Extra excluded directory name. Repeatable or comma-separated.
--case-insensitive-paths / --no-case-insensitive-paths
Case-fold relative path keys. Default: true.
--blank-comments / --no-blank-comments
Blank comments before scanning changed files. Default: true.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Include scanner diagnostic fields in updated file indexes.
Compatibility alias: --emit-debug-file-indexes.
Default: false.
--dry-run Show added/modified/deleted files without writing.
--jobs N Worker processes for added/modified files.
--force Reindex all current files.
--known-files-only Only consider files already present in manifest.json.
--changed-file PATH Known changed file. Repeatable. Used by watchers to avoid
hashing unchanged files.
--progress / --no-progress Show progress on stderr. Default: true.
--print-summary-json Print summary JSON.
--summary-json-file PATH Write summary JSON while keeping normal console output.
Poll source files and run incremental updates after changes settle.
--root PATH Project/source root. Default: current directory.
--index-root PATH Index root. Default: <root>/.mcp-cpp-project-indexer.
--indexer-root PATH Directory containing the indexer scripts.
--extensions EXT [EXT ...] Source extensions for snapshot scanning.
--include-extensionless-headers / --no-include-extensionless-headers
Also discover extensionless files that look like
C/C++ headers. Uses a conservative first-lines
heuristic. Default: false.
--git-ignore / --no-git-ignore Filter discovered files through git check-ignore
when available. Default: true.
--exclude-dir NAME Extra excluded directory name. Repeatable or comma-separated.
--case-insensitive-paths / --no-case-insensitive-paths
Case-fold relative path keys. Default: true.
--jobs N Worker process count for update actions.
--poll-interval SECONDS Poll interval. Default: 1.0.
--debounce SECONDS Wait for changes to settle. Default: 1.5.
--module-map / --no-module-map Rebuild module_map.json after real index changes. Default: true.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Pass diagnostic emission to watcher-triggered updates.
Compatibility alias: --emit-debug-file-indexes.
Default: false.
Build module relationship metadata from the project index.
--index-root PATH Project index root. Default: ./.mcp-cpp-project-indexer.
--output PATH Output JSON path. Default: <index-root>/module_map.json.
--print-summary-json Print compact summary JSON.
Dump readable module tree/import views.
--index-root PATH Project index root. Default: ./.mcp-cpp-project-indexer.
--imports MODULE Dump import tree for one C++20 module name.
--max-depth N Maximum tree depth. For --imports, default is 4.
--max-files N Maximum file paths per module leaf. Default: 1.
--output PATH Optional output text file.
Run the MCP stdio server.
--project-root PATH Project root used for read_range/read_symbol.
--index-root PATH Directory containing the generated index.
--watch-index / --no-watch-index
Start server-managed background watcher. Default: false.
--watch-poll-interval SECONDS Watcher poll interval. Default: 1.0.
--watch-debounce SECONDS Watcher debounce delay. Default: 1.5.
--watch-jobs N Worker process count for watcher update actions.
--watch-module-map / --no-watch-module-map
Rebuild module_map.json after watcher updates. Default: true.
--watch-emit-diagnostic-file-indexes / --no-watch-emit-diagnostic-file-indexes
Pass diagnostic emission to server watcher updates.
Compatibility alias: --watch-emit-debug-file-indexes.
Default: false.
--watch-include-extensionless-headers / --no-watch-include-extensionless-headers
Let the server watcher discover extensionless files
that look like C/C++ headers. Default: false.
--watch-git-ignore / --no-watch-git-ignore
Filter watcher discovery through git check-ignore
when available. Default: true.
--transport stdio|http Transport mode. Default: stdio.
--http-host HOST HTTP bind host for --transport http. Default: 127.0.0.1.
--http-port PORT HTTP bind port for --transport http. Default: 8765.
Terminal control center for build/update/watch/server workflows.
--root PATH C++ project root. Default: current directory.
--index-root PATH Index root. Default: <root>/.mcp-cpp-project-indexer.
--indexer-root PATH Directory containing indexer scripts.
--jobs N Worker process count for launched commands.
--http-url URL HTTP server base URL for live status.
Default: http://127.0.0.1:8765.
--enable-management-api / --no-enable-management-api
Enable the TUI's HTTP + management launch mode.
--management-token TOKEN Management API token used when launching
HTTP + management.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Initial diagnostic file section setting.
Optional Textual terminal UI with mouse support and live status panels. Requires
pip install -r requirements-ui.txt.
--root PATH C++ project root. Default: current directory.
--index-root PATH Index root. Default: <root>/.mcp-cpp-project-indexer.
--indexer-root PATH Directory containing indexer scripts.
--jobs N Worker process count for launched commands.
--http-url URL HTTP server base URL for live status.
Default: http://127.0.0.1:8765.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Initial diagnostic file section setting.
Legacy interactive menu and non-interactive command wrapper. Prefer
indexer_control.py for normal operation.
--root PATH C++ project root. Default: current directory.
--index-root PATH Index root. Default: <root>/.mcp-cpp-project-indexer.
--indexer-root PATH Directory containing the indexer scripts.
--jobs N Worker process count for build/update actions.
--emit-diagnostic-file-indexes / --no-emit-diagnostic-file-indexes
Initial menu switch state for scanner diagnostic file sections.
Compatibility alias: --emit-debug-file-indexes.
Default: false.
--action NAME Run one action without the menu.
Supported --action values:
build-index
build-module-map
build-all
update-dry-run
update-known-dry-run
update-index
update-known
update-all
update-known-all
watch
summary
diagnostics
dump-module-tree
lmstudio-config
server
server-watch
server-http-watch
get_project_summary
get_index_fingerprint
get_file_fingerprint(file)
get_symbol_fingerprint(symbolId)
get_data_fingerprint(dataId)
validate_fingerprints(items)
The server exposes an index stateFingerprint: a cheap server-side fingerprint
over the current index artifacts (manifest.json, index.sqlite,
module/diagnostic/update files). It changes when the loaded index state changes,
so relay/orchestrator layers can mark older tool evidence as stale after a
rebuild, update, or cache reload.
Every MCP tool result carries the fingerprint in result _meta:
{
"_meta": {
"stateFingerprint": "..."
}
}The same value is also exposed by get_project_summary.stateFingerprint and the
HTTP status endpoint under index.stateFingerprint.
For fine-grained evidence reuse, prefer the fingerprint tools. The global index fingerprint is only a warning signal. File, symbol, and data fingerprints let a relay validate active facts without dropping the whole evidence cache after every update:
fileFingerprint =
hash(fileId + relativePath + contentHash + line/token counts)
symbol/data fingerprint =
hash(id + fileFingerprint + startLine + endLine + signature)
validate_fingerprints is the batch entry point for relay/orchestrator layers.
It accepts up to 500 file/symbol/data items and returns compact validity and
fingerprint metadata only. These calls do not read or return source text.
These tools are exposed only when git.exe is available and project-root is
inside a worktree. They are read-only and do not expose a generic command runner.
list_changed_files
list_recent_revisions
get_revision_summary
get_file_change_hunks
resolve_hunk_to_indexed_range
Use them for current changes, recent revisions, hunk inspection, review of modified files, and commit-message suggestions.
Use resolve_hunk_to_indexed_range(file, line|startLine/endLine) when hunk
metadata gives a changed new-line range and the relay needs a valid
symbolId/dataId routing target before reading source. It returns indexed
symbol/data ranges that contain, overlap, or are nearest to the changed range.
This is metadata-only and does not interpret the diff or claim correctness.
When the user says they fixed, changed, saved, updated, committed, or wants current work reviewed, the agent should check these tools before normal source navigation. In watcher setups, the index may already be current, and the change tools provide the cheapest route to the affected file/symbol ranges.
get_file_change_hunks can include indexedRanges, which are intersections
between changed hunk line ranges and indexed symbol/data ranges. These are
routing hints only. Read the relevant source with read_symbol or read_range
before making implementation claims.
For low-token review routing, call get_file_change_hunks with
includeIndexedRangeSummary:true, includeIndexedRanges:false, and
includeSource:false. The returned summaryByIndexedRange groups changed
hunks by affected indexed symbol/data range without repeating every per-hunk
intersection.
After selecting an affected range, pass symbolId or dataId back to
get_file_change_hunks to return only hunks intersecting that symbol/data
range. This is still line-range filtering, not semantic analysis.
find_symbol(query)
find_declaration(query)
find_symbols_glob(pattern)
read_symbol(symbolId)
read_range(file, startLine, endLine)
read_range(file, line, beforeLines, afterLines)
get_nearest_symbol_for_line(file, line)
list_file_symbols(file)
get_nearest_symbol_for_line maps a file/line from diagnostics, hunks, build
output, Visual Studio, or IDA notes to containing or nearest indexed symbol/data
ranges. It is metadata-only; read the selected source range before making
behavior claims.
read_symbol accepts optional startOffset/endOffset or absolute
startLine/endLine to read only a slice of a large symbol body.
read_range accepts either explicit startLine/endLine or a compact
around-line form with line, beforeLines, and afterLines. The around-line
form is intended for diagnostics, hunks, search matches, Visual Studio handoff,
and IDA notes where the caller has one relevant source line.
find_symbol searches symbol metadata only:
- exact
shortName - exact
qualifiedName/ aliases - fallback substring over
shortName,qualifiedName, andsignature
Optional routing controls:
compact: return only compact routing fieldsresponseFormat:prettyorminifiedJSON for metadata responsesomitNulls: omit null fields from metadata responsesomitEmpty: omit empty arrays/objects from metadata responsessymbolTypes: filter by indexed symbol kind, e.g.method,function,type_aliascontainer: filter to symbols contained by a class/struct/namespace name or suffixfile: filter to one fileId or project-relative pathfilePattern: filter by project-relative path globexactOnly: return only exact short-name or qualified-name matches, including case-insensitive exact matcheshideNamespaces: hide namespace reopening symbols from navigation results
Use container only for the actual lexical/index container where the symbol is
declared. Do not use a class container for helper free functions just because
that class calls them. If a find_symbol query with container returns no
result, retry the exact symbol name without container before falling back to
search_source.
Each returned item includes matchKind to describe why it matched. Strong
matches such as exact_qualified_name and exact_short_name are usually best
for routing. Substring, signature, and metadata matches are weaker candidates
that should be disambiguated before reading source.
Do not combine file and filePattern. These controls are locator filters;
they do not read source code and do not resolve overloads semantically.
Response packing options are exposed only on metadata/routing tools. Source tools keep their line-numbered source output unchanged.
list_file_symbols can also return smaller file-level symbol candidate lists when the file is already known:
compact: return only compact routing fieldssymbolTypes: filter by indexed symbol kindcontainer: filter to symbols contained by a class/struct/namespace name or suffixhideNamespaces: hide namespace reopening symbolslimit: bound the result size
This is a locator filter only. It does not resolve inheritance, overloads, or type semantics.
When a file is already known and both class members and local helper functions
may matter, prefer list_file_symbols(file, compact:true, hideNamespaces:true)
before guessing a container filter. Anonymous-namespace helpers are still
normal indexed functions; they may have the surrounding named namespace as their
container, not the class that calls them.
search_source accepts symbolId to search only inside one indexed symbol
range. This is the preferred way to do a lexical check inside an already-located
function or method. It is still lexical source search, not semantic
call/reference resolution.
Canonical argument:
{ "query": "Editor::_OnScroll" }name may be accepted as a compatibility alias, but query is canonical.
find_files(pattern)
list_file_includes(file)
get_file_structure(file)
Glob over project-relative paths only. This is not source grep.
list_file_includes returns the lexical #include directives for one file.
It is intended for classic include-based C++ codebases such as Chromium:
{
"file": "chrome/app/chrome_main.cc",
"compact": true,
"includeResolved": true
}Include metadata is routing evidence only. The indexer records the source line,
target text, include kind (quote, angle, or macro), and best-effort
project-relative resolution when the included file is directly visible. It does
not evaluate #if/#ifdef, compiler include directories, generated headers, or
macro expansion.
get_file_structure returns a metadata-only table of contents for one file. For large files, prefer includeOutline:false first. Use includeIncludes:true when include metadata is needed. Use symbolTypes, dataKinds, hideNamespaces, outlineLimit, and compactOutline:true to keep responses small. After identifying a relevant outline item, read the source with read_symbol or read_range before making implementation claims.
When file indexes were built with --emit-diagnostic-file-indexes,
get_file_structure can also include optional parser/indexer diagnostic
sections:
{
"file": "...",
"includeOutline": false,
"includeIndexerDiagnostics": true,
"diagnosticKinds": ["structuralEvents", "scopeIntervals", "functionBodyRanges"],
"diagnosticStartLine": 120,
"diagnosticEndLine": 180,
"compactDiagnostics": true,
"diagnosticLimit": 100
}Use this only to investigate parser diagnostics, missing symbols, suspicious source ranges, or unexpected scope/function-body detection. Indexer diagnostic output is scanner evidence, not implementation behavior.
Prompt wording matters here. The phrase "debug information" is too broad for
many AI agents and can be confused with C++ debug-build code such as DEBUG,
_DEBUG, #ifdef DEBUG, Visual Studio debugging, or logging branches. For
indexer scanner data, ask for indexer/parser diagnostics explicitly.
Good prompts:
Show me which indexer diagnostics are present.
Show me the parser/indexer diagnostics from the index.
Show me indexerDiagnostics.diagnostics from get_file_structure(includeIndexerDiagnostics:true) for file X.
Which indexerDiagnostics.diagnostics are available in the index for file X?
Show me the source location for this indexer diagnostic.
Avoid ambiguous prompts such as:
Show me debug information.
Find the debug code.
Where is DEBUG used?
The intended workflow is:
get_file_structure(includeIndexerDiagnostics:true, compactDiagnostics:true)
-> inspect indexerDiagnostics.diagnostics first
-> read_range around the reported line
-> optionally get_nearest_symbol_for_line for the containing symbol
Examples:
*Editor*
*/TextEditor/*.ixx
*/Shell/Browser/*
find_module(moduleName)
list_module_files(moduleName)
search_modules(pattern)
get_module_map_summary
get_module_info(moduleName)
list_module_imports(moduleName)
list_module_imported_by(moduleName)
get_module_tree(maxDepth)
Use C++20 module syntax:
Example.Module:Partition
uiframework.Elements:ElementImpl
Do not pass C++ namespace syntax to module tools:
Example::Namespace
UIFramework::Elements
For namespaces/classes/functions, use find_symbol or find_symbols_glob.
Direction matters:
What does module X import? -> list_module_imports(X) or get_module_info(X)
Who imports/consumes module X? -> list_module_imported_by(X) or get_module_info(X)
Do not answer reverse-import questions by searching source text first. The
module map already stores importedBy metadata. Use source reads only when the
caller asks to inspect the actual import line or when metadata looks suspicious.
Module metadata tools accept compact:true where useful:
find_module/list_module_files: compact module/file routing fieldsget_module_info: compact files, imports, and imported-by metadatalist_module_imports: compact outgoing import recordslist_module_imported_by: compact importing-module records
find_data(query)
list_type_members(container)
read_data(dataId)
resolve_code_entity(query, file?, line?, container?)
Use find_data for fields, globals, namespace constants, enum values, variable
templates, and concepts. Use list_type_members when the containing type or
namespace is already known. Both tools are metadata-only and accept
compact:true for smaller routing output.
Use resolve_code_entity when a source range contains an identifier and the AI
needs to decide whether it is probably a field/data declaration, callable
symbol, type symbol, or ambiguous candidate before choosing the next read tool.
It accepts optional file, line, and container context and returns ranked
metadata candidates plus a recommended next tool such as read_data or
read_symbol.
resolve_code_entity is still orientation only. It does not perform compiler
name lookup, overload resolution, macro expansion, C++ type resolution, or
semantic reference resolution. Source-level claims still require
read_data, read_symbol, or read_range.
typeText is the original source text for the declaration type. It is not a
compiler-resolved type. Treat it as a hint for further symbol/source lookup, not
as proof of semantic type identity.
The model should follow these rules:
Use mcp-cpp-project-indexer as a deterministic source-range locator.
Read source before making implementation claims.
Use query as the canonical argument for symbol lookup tools.
Do not ask for analyze_symbol.
Do not ask for a precomputed call graph.
Do not treat module metadata as implementation behavior.
Large MCP tool sets work best when the client keeps the active tools relevant to the current task. If every tool is exposed for every request, some models may over-explore, repeat similar queries, or choose broader tools than necessary. This is normal behavior for general-purpose models and does not indicate a problem with the index data.
In typical use, start with compact metadata queries and read source ranges only when implementation evidence is needed.
Correct calls:
find_symbol({"query": "Editor::_OnScroll"})
find_declaration({"query": "OnNotifyReflect"})
Avoid:
find_symbol({"name": "Editor::_OnScroll"})
The full system prompt template is available in prompt_template.md.
User:
Find the declaration and definition of OnNotifyReflect.
AI workflow:
find_symbol({"query": "OnNotifyReflect"})
read_symbol(symbolId for declaration)
read_symbol(symbolId for definition)
User:
Show me Editor::_OnScroll.
AI workflow:
find_symbol({"query": "Editor::_OnScroll"})
read_symbol(symbolId)
Then the AI follows only relevant project-local calls:
GetHWND -> find_symbol/read_symbol
SendMessageW -> external Win32 API, do not query project index
MAKEWPARAM -> external Win32 macro, do not query project index
Workflow:
1. get_module_info or list_module_imports/list_module_imported_by
2. use relativePath and sourceLine from import metadata
3. list_file_symbols(relativePath)
4. choose the likely entry point from symbol names/signatures
5. read_symbol(candidate)
6. inspect visible source usage
7. follow more project symbols only if needed
Do not use find_symbols_glob as a substitute for source usage search. It searches symbol metadata, not callsites.
symbolId -> fileId -> startLine/endLine -> original source
Runtime index stores routing facts. Scanner diagnostic data is opt-in with
--emit-diagnostics or --emit-diagnostic-file-indexes.
index.sqlite -> symbol/data search
module_map.json -> module browsing
manifest.json -> file browsing
Do not hide real parser/source problems just to reach zero diagnostics.
No call graph. No reference graph. No macro expansion. No analyzer tool. The AI recursively explores source on demand.
cd <project-root>
python <indexer-root>\build_project_index.py
python <indexer-root>\build_module_map.py `
--index-root .\.mcp-cpp-project-indexerThen restart the MCP server / LM Studio.
This project is licensed under the Apache License 2.0.
SPDX-License-Identifier: Apache-2.0
Add the full Apache-2.0 license text in the repository as:
LICENSE
Optional source-file header:
# SPDX-License-Identifier: Apache-2.0Copyright (c) 2026 Mike Walter
Project name:
mcp-cpp-project-indexer
This project was shaped through a multi-model AI-assisted development workflow involving ChatGPT, Gemini, and DeepSeek. The goal was not to let one model blindly invent a broad tool, but to use multiple models and real production feedback to argue about scope, constraints, and workflow friction.
-
Architecture and scope debate ChatGPT and Gemini were used to debate the boundary of the tool: what it should do, what it should avoid, and why it should stay a lightweight routing index instead of becoming a compiler or
clangdreplacement. -
Incremental implementation The codebase was implemented incrementally with an AI coding agent, with each functional change tested against real C++20 module-heavy projects.
-
Workflow structure analysis and review DeepSeek was used as a dedicated workflow-analysis partner during production use inside a 7,000-file codebase. It evaluated tool-call sequences, parameter choices, output size, ranking quality, navigation friction, and context hygiene, then fed back concrete changes to reduce calls and improve token efficiency.
-
Iterative refinement Feedback from those sessions drove focused improvements: compact outputs, symbol/file/container filters, change hunk routing, watcher reloads, around-line reads, and tighter prompt rules.
The result is a tool built with AI for AI-assisted code navigation, optimized around one strict principle: keep the indexer honest and small, and let the AI reason only from source ranges it explicitly reads.
After MCP integration, test:
- get_project_summary
- Show all modules under Example.Core.Direct2D.
- Which modules import Example.Shell.Browser:Impl?
- Find the declaration and definition of ExampleClass::OnEvent.
- Read all overloads of GetHandler.
- Which file defines Example.Elements:ElementImpl?
- Find the definition of ExampleClass::operator=.
- List all symbols in Example/Core/Renderer.cpp.
- Which modules does Example.UI:ToggleSwitch import?
- Show the module tree for Example.Core (max depth 3).
- Find all files matching ExampleDialog*.
- Read the first 20 lines of Example/Core/Main.ixx.
- Find the declaration of ExampleClass::ExampleClass (constructor).
- Find all symbols matching Paint in Example/Core/Renderer.cpp.
- Show which modules are imported by Example.Shell:Impl.
- Get the module map summary.
- Find the definition of ExampleClass::OnKeyDown.
- Read the range of ExampleClass::OnPaint (lines 150-200).
- Find all files in module Example.UI:Controls.
- Show the import tree for Example.UI:ToggleSwitch.
- Find the symbol ExampleNamespace::ExampleClass::Method.
- List all modules that import Example.Core.Service.
- Find the declaration of template function ExampleClass::Create.
- Read the symbol for ExampleClass::OnNotify.
- Find all files under Example/UI/Controls/.
- Show the module info for Example.Core.Direct2D.Renderer.
- Find the definition of ExampleClass::Initialize.
- List all symbols in Example/Core/Service.cpp that are functions.
- Find which modules are part of Example.UI.
- Read the source of ExampleClass::HandleInput (lines 50-120).
- Find the file that defines Example.Core:ModuleImpl.
- Show all imports of Example.UI:Controls.Button.
- Find the declaration of enum ExampleClass::State.
- Read the symbol for ExampleClass::GetSize.
- Find all files with .ixx extension under Example/.
- Show the module tree for Example.Shell (max depth 2).
- Find the definition of ExampleClass::_UpdateLayout.
- List all symbols in Example/Core/Utils.cpp.
- Find which modules import Example.UI:Controls.
- Read the range of ExampleClass::Paint (lines 200-250).
- Find the declaration of ExampleClass::~ExampleClass (destructor).
- Show the module info for Example.Shell.Browser:Impl.
- Find all symbols matching Handler in Example/Core/.
- Read the symbol for ExampleClass::OnMouseMove.
- Find the file that defines Example.UI:Controls.Button.
- List all modules imported by Example.Core.Direct2D.
- Find the definition of ExampleClass::SetValue.
- Show the module tree for Example.UI (max depth 4).
- Find all files under Example/Shell/Browser/.
- Read the first 30 lines of Example/Core/Service.ixx.
- Find the declaration of ExampleClass::GetAccessibleImpl.
- List all symbols in Example/UI/Controls/Button.cpp.
- Find which modules are imported by Example.Shell.Browser:Impl.
- Show the module info for Example.Elements:ElementImpl.
- Find the definition of ExampleClass::OnPropertyChanged.
- Read the symbol for ExampleClass::GetState.
- Find all files matching Service under Example/Core/.
- Show the module tree for Example.Elements (max depth 3).
- Find the declaration of ExampleClass::Register.
- List all symbols in Example/Shell/Browser/Impl.cpp.
- Find which modules import Example.Core.Direct2D.Renderer.
- Read the range of ExampleClass::OnInput (lines 100-150).
- Find the definition of ExampleClass::_SelfLayoutDoLayout.
- Show the module info for Example.UI:Controls.Dialog.
- Find all symbols matching Layout in Example/Core/.
- Read the symbol for ExampleClass::GetSliderSize.
- Find the file that defines Example.Shell:Impl.
- List all modules that are part of Example.Core.
- Find the declaration of ExampleClass::DefaultAction.
- Read the first 40 lines of Example/UI/Controls/Dialog.ixx.
- Find the definition of ExampleClass::_FireClickEvent.
- Show the module tree for Example.Core.Direct2D (max depth 5).
- Find all files under Example/Elements/.
- Read the symbol for ExampleClass::GetCaptured.
- Find which modules import Example.Private.Helper.
- Show the module info for Example.UI:Controls.Dialog.
- Find the declaration of ExampleClass::SetCaptured.
- List all symbols in Example/Elements/ElementImpl.cpp.
- Find the definition of ExampleClass::OnInput.
- Read the range of ExampleClass::_UpdateLabel (lines 170-180).
- Find all files matching Element under Example/Elements/.
- Show the module tree for Example.Private (max depth 2).
- Find the declaration of ExampleClass::GetPressed.
- Read the symbol for ExampleClass::SetPressed.
- Find which modules import Example.Elements:ElementImpl.
- Show the module info for Example.Private.Helper.
- Find the definition of ExampleClass::OnPropertyChanged.
- List all symbols in Example/Core/Direct2D/Renderer.cpp.
- Find all files with .cpp extension under Example/UI/.
- Read the symbol for ExampleClass::GetSliderBorderStrokeWidth.
- Find the declaration of ExampleClass::SliderBackgroundColorProp.
- Show the module tree for Example (max depth 1).
- Find all modules that start with Example.UI.
- Read the range of ExampleClass::_SelfLayoutUpdateDesiredSize (lines 200-220).
- Find the definition of ExampleClass::SliderBorderColorProp.
- List all symbols in Example/Private/Helper.cpp.
- Find which modules import Example.UI:Controls.Dialog.
- Show the module info for Example.Elements:ElementWithTooltip.
- Find the declaration of ExampleClass::SliderForegroundColorProp.
- Read the symbol for ExampleClass::SliderSizeProp.
- Find all files under Example/Private/.
- Find the definition of ExampleClass::StateProp.
- List all symbols in Example/Core/Service.cpp.
- Find which modules are imported by Example.Elements:ElementImpl.
- Show the module tree for Example.UI.Controls (max depth 3).
- Find the declaration of ExampleClass::CapturedProp.
- Read the symbol for ExampleClass::PressedProp.
- Find all files matching Helper under Example/Private/.
- Find the definition of ExampleClass::ClassInfoI::GetClassInfoPtr.
- List all symbols in Example/UI/Controls/Dialog.cpp.
- Find which modules import Example.Private.TreeHelper.
- Show the module info for Example.UI:Controls.Button.
- Find the declaration of ExampleClass::ElementWithPromptValueI::ElementWithPromptValueProperties.
- Read the symbol for ExampleClass::Initialize (overload with 3 parameters).
- Find all files under Example/Core/Direct2D/.
- Find the definition of ExampleClass::_Initialize.
- List all symbols in Example/Elements/ElementWithTooltip.cpp.
- Find which modules import Example.Private.ValueHelper.
- Show the module tree for Example.Shell.Browser (max depth 4).
- Find the declaration of ExampleClass::Register.
- Read the symbol for ExampleClass::ToggleSwitch (constructor).
- Find all files matching Renderer under Example/Core/Direct2D/.
Expected behavior:
- the model uses tools
- it does not read whole files
- it does not confuse namespaces with modules
- it calls
read_symbolonly after symbol metadata lookup - it follows project calls recursively only when needed
After changing tool behavior, command-line switches, prompts, or documentation, run through MAINTENANCE_CHECKLIST.md. It covers the usual sync points: tool descriptions, prompt rules, README, menu/CLI flags, index data compatibility, smoke tests, relay assumptions, and commit prep.