Extracts string values from configuration files, data files, and various source code languages for i18n preparation and config review.
Also known as nolindnaidoo/string-le, string-le, string-le-mcp
- Stars
- 1
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- 2.5.1
- Last commit
- Oct 31, 2026
- License
- MIT
What is Strings-LE MCP?
Strings-LE extracts string values from configuration files, data files, and various source code languages for i18n preparation and config review.
Summary generated from the project's README
About Strings-LE MCP server
Extracting translatable text and configuration values from diverse file formats is what Strings-LE enables for developers preparing projects for internationalization. By using the extract_strings tool, an AI agent can pull string values from JSON, YAML, TOML, INI, and environment files, as well as ten different source code languages including Python, Rust, Go, and TypeScript. This allows for rapid i18n preparation and configuration reviews without manual searching.
This MCP server runs via stdio and requires no API keys or special environment variables. It is designed to be used locally with clients like Claude Code or Cursor by pointing them to the npx string-le-mcp command. The server processes the content provided by the client and returns the extracted values in document order, with a default limit of 500 results. It focuses on returning actual string values while ignoring keys, numbers, and booleans to ensure the output is clean and relevant for translation tasks.
What you can do with it
- 1Prepare files for internationalizationAn AI can flatten complex locale files into a simple list of translatable values.
- 2Review configuration settingsDevelopers can use an agent to quickly see every string value stored in a TOML or INI file.
- 3Mine data from large CSV filesYou can instruct an agent to extract specific columns or all cells from massive data files using streaming mode.
Written from the project's README. Every use case is checked against a quote from it.
How to install Strings-LE MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Strings-LE MCP in Claude Code
$ claude mcp add string-le -- npx -y string-le-mcpInstall Strings-LE MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"string-le": {
"command": "npx",
"args": [
"-y",
"string-le-mcp"
]
}
}
}Install Strings-LE MCP in VS Code
.vscode/mcp.json
{
"servers": {
"string-le": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"string-le-mcp"
]
}
}
}Install Strings-LE MCP in VS Code CLI
$ code --add-mcp '{"name":"string-le","type":"stdio","command":"npx","args":["-y","string-le-mcp"]}'Install Strings-LE MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"string-le": {
"command": "npx",
"args": [
"-y",
"string-le-mcp"
]
}
}
}Install Strings-LE MCP in Codex
$ codex mcp add string-le -- npx -y string-le-mcpInstall Strings-LE MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.string-le] command = "npx" args = ["-y", "string-le-mcp"]
Install Strings-LE MCP in Gemini CLI
$ gemini mcp add -s user string-le npx -- -y string-le-mcpInstall Strings-LE MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"string-le": {
"command": "npx",
"args": [
"-y",
"string-le-mcp"
]
}
}
}Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachablenolindnaidoo/string-le is public and reachablegithub.com/nolindnaidoo/string-le ↗Pass
- Package names this repository as its sourcestring-le-mcp names nolindnaidoo/string-le as sourcenpmjs.com/package/string-le-mcp ↗Pass
- Build provenance (Sigstore attestation)No build attestation published for this releaseNot checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-10-31github.com/nolindnaidoo/string-le ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from string-le-mcpPass
- No critical advisory in the package or its dependencies1 resolved dependencies checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubUseful? A star or rating is how other developers find it — ★ GitHub · ★ Open VSX · ★ Marketplace
What it does
Open a file, run String-LE: Extract Strings, and every string value in the document lands in a new editor — deduplicate and sort it from there. Works in VS Code and in VS Code–based editors like Cursor and VSCodium (installable from Open VSX).
- i18n prep — flatten locale files (JSON/YAML) into a clean list of translatable values
- Config review — see every string value in a TOML/INI/.env file at a glance
- CSV mining — pull one column, several, or all of them; stream very large files
Install
| Where | What you get | Install |
|---|---|---|
| VS Code | The same extraction, in your editor, on a keystroke | Marketplace |
| Cursor, VSCodium, Windsurf | The same extension | Open VSX |
| A terminal or a CI step | The same run over a whole tree, with exit codes | npx strings-le · npm or cargo install string-le · crates.io |
| Any MCP agent, via Node | extract_strings over stdio |
npx string-le-mcp · npm |
Use it from an AI agent
The same engine runs as an MCP server, so an agent can call it directly instead of you running a command.
| Editor | How |
|---|---|
| VS Code 1.101+ | Nothing to install — the extension registers extract_strings with agent mode |
| Claude Code | claude mcp add string-le -- npx -y string-le-mcp |
| Cursor, Windsurf, anything else | point it at npx string-le-mcp |
extract_strings(content, format?, filename?, dedupe?, multiline?, maxResults?)
Returns the values in document order, capped at 500 by default with meta.truncated. A format is optional — any unrecognised format falls back to quoted strings.
The server takes content and returns data — it reads no files and makes no network requests of its own. Published as string-le-mcp on npm and as io.github.nolindnaidoo/string-le in the MCP registry.
Most hosts read a JSON config. Add one entry:
{
"mcpServers": {
"string-le": {
"command": "npx",
"args": ["-y", "string-le-mcp"]
}
}
}
-y skips the install prompt on first run. Pin a version if you would rather not track releases — string-le-mcp@2.5.1.
Prefer not to go through npx on every launch? Install it once and point at the binary instead:
npm install -g string-le-mcp
{
"mcpServers": {
"string-le": { "command": "string-le-mcp" }
}
}
It speaks MCP over stdio and needs no environment variables, no API key and no configuration of its own. To check it before wiring it into anything:
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | npx -y string-le-mcp
That prints the tool list and exits — if you see extract_strings, the server works.
Supported formats
| Format | Language IDs | What gets extracted |
|---|---|---|
| JSON | json |
String values (parsed; keys, numbers, and booleans excluded) |
| YAML | yaml |
String values including unquoted plain scalars, block/folded scalars, and multi-document files |
| CSV | csv |
Cells, with optional header handling, column selection, and a streaming mode for large files |
| TOML | toml |
String values including multiline strings; dates and numbers are excluded as typed values |
| INI | ini |
Values (INI is untyped, so numeric-looking values are extracted as strings) |
| Environment | dotenv, env |
Values (export prefixes, quotes, and inline comments handled) |
| Python | python |
Triple-quoted docstrings as one string, f-strings, r/b/u prefixes |
| Rust | rust |
Raw strings r"…", r#"…"#, r##"…"## as one string with their inner quotes; byte strings |
| Go | go |
Backtick raw strings as one string, across lines |
| Shell | shellscript |
Heredocs (<<EOF, <<'EOF', <<-EOF) as one string |
| PHP, Ruby, Perl | php, ruby, perl |
Heredocs and nowdocs (<<<EOT, <<<'NOW', <<~EOS) as one string |
| C# | csharp |
Verbatim @"…" where "" is one quote; interpolated $"…" |
| JavaScript, TypeScript | javascript, typescript (+ the react ids) |
Template literals as one string, interpolation and nesting included |
| Markdown, anything else | markdown |
Fallback scan for "double", 'single', or `backtick` quoted strings on a single line |
Values are trimmed; empty values are dropped; keys are never extracted — one rule, shared by every extractor above. The fallback scan cannot see unquoted strings, which is why the parsed formats and the source languages get real readers; it reads a quoted run across lines when string-le.fallback.multiline is on (the MCP tool's multiline). Parse errors are silent unless string-le.showParseErrors is on.
Across a folder or a workspace
Extract reads the document you have open. A scan reads many files from disk and gives one report.
- The whole workspace: run
String-LE: Extract Strings from Workspacefrom the command palette. - One folder: right-click it in the Explorer and choose
Extract Strings from Folder, or runString-LE: Extract Strings from Folderand pick one.
A project writes the same string in many places, so the report is the distinct strings, the most widely used first, with how often each is written and where. Every string is in the table. Most are written once, and such a one is placed in its row. A repeated one gets a section below that lists each place:
# String-LE workspace report
`my-project` · 4 file(s) read · 3 distinct string(s), 7 occurrence(s) in 3 file(s)
| String | Occurrences | Files | Where |
|---|---|---|---|
| `Save changes` | 4 | 3 | |
| `Cancel` | 2 | 2 | |
| `Press Enter` | 1 | 1 | `src/b.py` · **2:8** |
## `Save changes` (4)
- `i18n/en.json` · **2:11**
- `src/a.ts` · **1:11**, **3:11**
- `src/b.py` · **1:9**
## `Cancel` (2)
- `i18n/en.json` · **3:13**
- `src/a.ts` · **2:11**
## Could not be read (1)
- `bad.json`: Invalid JSON: Expected value after colon in object property on line 1 column 6
Positions follow string-le.showPositions. With it off, each line is the file and how many times the string is in it: src/a.ts (2). The copy on the clipboard follows string-le.clipboardIncludesPositions, as it does for Extract.
A file its format reader could not parse is listed with the reason, never passed over. A string longer than 120 characters, or one with a line break, is shown as one cut line. Two strings are still compared whole.
What a scan reads. Files come from disk, so an unsaved edit is not seen. A file over the safety size, or one that is not UTF-8 text, is left unread. It stops at 5,000 files or 10,000 listed occurrences. The report ends with a line for each thing it left out, so a short report is never mistaken for a clean project.
What it skips, and how to change that. Three switches are on by default, and each can be turned off on its own in Settings:
| Switch | Skips |
|---|---|
scanUseDefaultExcludes |
Dependency folders, build output, tool caches and lockfiles. The full list is below |
scanRespectGitignore |
Whatever the project's .gitignore files skip |
scanSkipBinaryFiles |
Images, fonts, archives and other files that are not text |
Two lists adjust the result without turning a switch off. To skip more, add a pattern to scanExcludes. To read something a switch would skip, add it to scanAlwaysInclude:
{
// Also skip the test fixtures.
"string-le.workspace.scanExcludes": ["**/fixtures/**"],
// Read the vendored code, though the built-in list skips it.
"string-le.workspace.scanAlwaysInclude": ["**/vendor/**"]
}
String-LE: Open Settings opens all of these in the Settings editor.
Folders, wherever they appear:
.git, .hg, .svn, node_modules, bower_components, jspm_packages, .pnpm-store, .yarn, vendor, site-packages, Pods, Carthage, dist, build, out, target, _build, _site, dist-newstyle, zig-out, storybook-static, cdk.out, DerivedData, CMakeFiles, .next, .nuxt, .output, .svelte-kit, .angular, .astro, .docusaurus, .vuepress, .expo, .turbo, .parcel-cache, .cache, .sass-cache, .jekyll-cache, .dart_tool, .pub-cache, .gradle, .kotlin, .cxx, .externalNativeBuild, captures, ephemeral, .symlinks, .swiftpm, .build, .bundle, .stack-work, .zig-cache, .godot, elm-stuff, .vercel, .netlify, .serverless, .aws-sam, .terraform, .venv, venv, __pycache__, .tox, .nox, .mypy_cache, .pytest_cache, .ruff_cache, .ipynb_checkpoints, .eggs, coverage, htmlcov, .nyc_output, .vscode-test, .idea, .vs, xcuserdata, *.egg-info
Files, wherever they appear:
*.min.js, *.min.css, *.map, *.snap, *.lock, package-lock.json, pnpm-lock.yaml, npm-shrinkwrap.json, go.sum, *.pbxproj, *.iml, local.properties, output-metadata.json, .flutter-plugins, .flutter-plugins-dependencies, .packages, Generated.xcconfig, flutter_export_environment.sh, GeneratedPluginRegistrant.*, fastlane/report.xml, fastlane/test_output/**, doc/api/**
Not on the list, because they are ordinary folders in many projects: bin, obj, tmp, logs, public, generated. A project that generates those ignores them in git, and the scan reads .gitignore.
The settings that shape a scan are under Settings.
The CLI
The same extraction runs from a terminal or a shell pipeline: a Rust CLI
in crate/, sharing one corpus with the extension —
crate/fixtures/ — so the two can never read a
document differently. Install it from npm with npm install -g strings-le (or
run it with npx strings-le) or from crates.io with cargo install string-le.
Both install the same command and are held to the same output byte for byte,
apart from the words a parser uses to refuse a broken document.
string-le . # every string in the tree; JSON on stdout when piped
string-le --values src/ # just the values, one per line
string-le --dedupe --values . # each distinct string once
string-le mcp # the same extraction over MCP on stdio
# the point of the whole thing:
string-le --values --dedupe src/ | sort > after.txt
diff before.txt after.txt # what changed in the copy this release
Install it from npm (strings-le) or from crates.io (string-le). Both install the same string-le command.
npm install -g strings-le # or run it once: npx strings-le
cargo install string-le
The reader is not the author. The extension answers for the buffer you have open. Someone still has to read every user-visible string before a release — a QA lead, a compliance reviewer, a localisation owner — and they do not have the editor open, and several of them cannot be handed a checkout at all. The CLI puts the whole repository into one file they can read.
Source files are the main event there. A .ts or .py file is where
the user-facing copy lives, so each language is read by its own literal
syntax and anything still unrecognised falls through to quoted-string
extraction rather than being refused.
Exit codes follow grep — 0 strings found, 1 none found, 2 the question was malformed — so finding nothing is an answer rather than an error.
Commands
| Command | Description |
|---|---|
String-LE: Extract Strings |
Extract all string values from the active document |
String-LE: Extract Strings from Workspace |
The distinct strings in every file in the workspace, and where each one is |
String-LE: Extract Strings from Folder |
The same for one folder. Also on a folder in the Explorer |
String-LE: Deduplicate Strings |
Remove duplicate lines from the active document |
String-LE: Sort Strings |
Sort lines alphabetically or by length |
String-LE: Toggle CSV Streaming |
Enable/disable streaming for large CSV files |
String-LE: Open Settings |
Open String-LE settings |
String-LE: Help & Troubleshooting |
Built-in documentation |
No command is bound to a key by default. Give any of them one under Keyboard Shortcuts in the editor.
Settings
| Setting | Default | Description |
|---|---|---|
string-le.openResultsSideBySide |
true |
Open results beside the current editor |
string-le.postProcess.openInNewFile |
true |
Post-process commands write to a new file instead of editing in place |
string-le.showPositions |
false |
Show the line and column of each string |
string-le.copyToClipboardEnabled |
false |
Also copy results to the clipboard (disabled for CSV output) |
string-le.clipboardIncludesPositions |
false |
Include the line and column in that copy |
string-le.dedupeEnabled |
false |
Deduplicate results automatically after extraction |
string-le.sortEnabled |
false |
Sort results automatically after extraction |
string-le.sortMode |
off |
alpha-asc, alpha-desc, length-asc, length-desc |
string-le.csv.streamingEnabled |
false |
Stream CSV results into the editor incrementally |
string-le.showParseErrors |
false |
Show parse errors as notifications |
string-le.notificationsLevel |
silent |
all = every notification, important = warnings + errors, silent = errors only |
string-le.workspace.scanPatterns |
["**/*"] |
The files a folder or workspace scan reads |
string-le.workspace.scanUseDefaultExcludes |
true |
Skip dependency folders, build output, caches and lockfiles |
string-le.workspace.scanRespectGitignore |
true |
Skip what the project's .gitignore files skip |
string-le.workspace.scanSkipBinaryFiles |
true |
Skip images, fonts, archives and other files that are not text |
string-le.workspace.scanExcludes |
[] |
More files to skip, as glob patterns |
string-le.workspace.scanAlwaysInclude |
[] |
Files to read even when one of the three above would skip them |
string-le.workspace.scanMaxFiles |
5000 |
The most files one scan reads |
string-le.workspace.scanMaxResults |
10000 |
The most occurrences one scan lists before it stops reading |
string-le.safety.enabled |
true |
Guardrails for very large files/outputs |
string-le.safety.fileSizeWarnBytes |
1000000 |
Warn before extracting above this file size |
string-le.safety.largeOutputLinesThreshold |
50000 |
Offer Open/Copy/Cancel above this result count |
string-le.safety.manyDocumentsThreshold |
8 |
Confirm before opening this many result documents (CSV multi-column) |
string-le.statusBar.enabled |
true |
Show the status bar item |
string-le.telemetryEnabled |
false |
Local-only event log (see Privacy) |
Languages
Twelve languages besides English:
German · Spanish · French · Indonesian · Italian · Japanese · Korean · Portuguese (Brazil) · Russian · Ukrainian · Vietnamese · Chinese (Simplified)
Both halves are covered — the manifest (command titles, setting names and descriptions) and everything shown while the extension runs (notifications, the status bar, quick-picks and prompts). The extension follows VS Code's display language, so it matches whatever the editor is already set to; no setting of its own.
Privacy & security
- No network access. The extension never sends data anywhere. The
telemetryEnabledsetting only writes events to a local Output Channel you can inspect (String-LE). - The MCP server holds the same line. It takes content as an argument and returns data: no filesystem access, no network calls, no telemetry. Your agent already has file-read tools, so duplicating them inside the server would add a path-traversal surface for no capability.
check:mcp-bundlefails the build if the server ever imports something that could reach either. - Error notifications redact home directories and credential-shaped fragments.
- One rating prompt, at most twice. On the 3rd successful use the extension asks once whether you would rate it, and once more on the 20th if you chose Later or dismissed it. Don't Ask Again ends it. Setting
notificationsLeveltoimportantorsilentyourself turns it off. The counts are kept in VS Code's extension storage and nothing is sent anywhere; Rate opens the listing you installed from — the VS Code Marketplace or Open VSX — in your browser.
Documentation
| What | Where |
|---|---|
| What the tool is allowed to say — scope, output contract, refusals, non-goals | crate/SPEC.md |
| How the extension is built and held together — architecture, invariants, toolchain, release | AGENTS.md |
| How the CLI is built and held together | crate/AGENTS.md |
| What changed | CHANGELOG.md · crate/CHANGELOG.md |
| The tool's page, and the other fifteen | letools.dev/tools/string-le |
Performance
| Input | Size | Found | Time | Rate | Scan speed |
|---|---|---|---|---|---|
| JSON locale file | 1.77 MB | 40,000 | 9.53 ms | 4,198,116/sec | 185.9 MB/s |
| YAML locale file | 1.01 MB | 30,000 | 13.33 ms | 2,250,169/sec | 75.7 MB/s |
| CSV strings | 1.35 MB | 80,002 | 28.29 ms | 2,827,438/sec | 47.8 MB/s |
Median of 7 runs after warmup, on Apple M5 Pro, 24 GB RAM, Node 24.3.0. Inputs are generated
by scripts/benchmark.ts rather than checked in, so the sizes above are
exactly what was measured. Reproduce with bun run benchmark.
These are machine-specific and are not asserted in CI — a benchmark that gates a build only tells you how busy the runner was.
Testing
| Metric | Coverage |
|---|---|
| Statements | 81.58% |
| Branches | 72.88% |
| Functions | 89.43% |
| Lines | 83.15% |
533 test cases across 43 files, plus an integration suite that runs
in a real VS Code extension host and an end-to-end test that installs the
built .vsix into a clean profile.
Generated from a real run — coverage/coverage-summary.json and
coverage/test-results.json — by scripts/coverage-readme.js, and
bun run coverage:readme:check says whether it has drifted since. Reproduce
with bun run test:coverage, and the case count is the one vitest prints.
Strings-LE MCP questions
How do I add Strings-LE MCP to Claude Code?
Run claude mcp add string-le -- npx -y string-le-mcp in your terminal. Add --scope user to make it available in every project.
How do I add Strings-LE MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What formats are supported for string extraction?
It supports JSON, YAML, CSV, TOML, INI, Environment files, and several source languages including Python, Rust, Go, Shell, PHP, Ruby, Perl, C#, and JavaScript/TypeScript.
Does the MCP server read my local files or make network requests?
No, the server only processes the content provided to it; it does not read files or make network requests on its own.
How do I add Strings-LE to Claude Code?
You can add it using the command: claude mcp add string-le -- npx -y string-le-mcp.
Can I use this with Cursor or Windsurf?
Yes, you can use it by pointing those editors to npx string-le-mcp.