Bragproof

← Reports

@groundtruth-mcp/gt-mcp

Starts with no setup8.0.1npm

Installed, started with no arguments, no environment and no input, and listed its tools. Run 2026-11-16.

Needs to start: nothing
Contacts while installing and starting: nothing

14tools listed
0hosts contacted
40files written
0.3 sto start
4dependencies

What it contacted

Nothing besides the npm registry it was installed from. Watched from outside its sandbox while it was installed and started with no input. Its tools may still contact things when called.

Starting it

npx -y -p @groundtruth-mcp/gt-mcp@8.0.1 gt-mcp

Installing a package runs its code. Use a sandbox for anything you have not read.

What its tools tell an agent

In the tool descriptions, nothing that talks about the user, other tools or private files, or uses hidden characters.

Found by simple pattern rules in each tool's description (up to 3,000 characters), quoted as written. Parameter descriptions are not scanned. A clean result is not a safety review.

Tools

gt_dispatch query, projectPath
Route a raw user request ("use gt", "use gt for react", "migrate next from 15 to 16") to the right gt_* tool. Returns {tool, args, reason, confidence}. Skip it when the tool is already clear.
gt_resolve_library libraryName, query
Find a library's GroundTruth ID ('vercel/next.js', 'npm:express') and its docs, llms.txt and GitHub URLs, from the registry, npm, PyPI, crates.io, Go and GitHub. Only for ambiguous or unknown names: other tools take plain names. 2-20 names: gt_batch_resolve. Registry is Elastic-2.0: look up named libraries only, never enumerate.
gt_get_docs libraryId, topic, version, tokens, projectPath
Reference docs for one library, filtered to a topic. libraryId: a plain name ('nextjs'), GroundTruth ID, npm:/pypi: package, or a docs URL. Curated guidance: gt_best_practices. Code-only snippets: gt_snippets.
gt_best_practices libraryId, topic, version, tokens
Current best practices for one library: patterns, performance, security, testing, migration. libraryId: a plain name ('nextjs'), GroundTruth ID or npm:/pypi: package. Prefer over gt_search when one library is named.
gt_auto_scan projectPath, topic, tokensPerLib
Detect the project's dependencies (package.json, pyproject.toml, requirements.txt, Cargo.toml, go.mod, ...) and fetch best practices for each. Issues in the project's own code: gt_audit.
9 more tools
gt_search query, tokens
Official docs and standards for a topic with no single library: web APIs, CSS, security, accessibility, protocols, infrastructure. One library named: gt_best_practices or gt_get_docs. Browser support: gt_compat. Real-world code: gt_examples.
gt_audit projectPath, categories, tokens, maxFiles
Scan the project's own source files for issues (file:line) and attach fixes from official docs. Use categories ["all"] for broad requests. Best practices for its dependencies: gt_auto_scan.
gt_changelog libraryId, version, tokens
Recent release notes for a library (GitHub Releases, CHANGELOG.md): what changed in a version. Step-by-step upgrade: gt_migration.
gt_compat feature, environments, tokens
Browser and runtime support for a web API, CSS feature or JS syntax ('container queries', 'Array.at()'), from MDN and caniuse. Takes a feature, not a library.
gt_compare libraries, criteria, tokens
Compare 2-3 libraries side by side from their live docs. Pass names (['prisma', 'drizzle-orm']), not IDs. One library: gt_get_docs.
gt_examples library, pattern, language, maxResults
Real-world usage of a library or pattern from public GitHub code, with repo links (needs GT_GITHUB_TOKEN, else official-docs examples). Snippets from the library's own docs: gt_snippets.
gt_migration libraryId, fromVersion, toVersion, tokens
Upgrade guide between two versions of a library: breaking changes and required code changes. Release notes only: gt_changelog.
gt_batch_resolve libraryNames
Resolve 1-20 library names to IDs and docs URLs in one call, registry only. One name with npm/PyPI fallback: gt_resolve_library. Project dependencies: gt_auto_scan. Registry is Elastic-2.0: look up named libraries only, never enumerate.
gt_snippets libraryId, topic, version, language, maxSnippets, refresh, projectPath
Ranked code snippets (title, language, code, source URL) from a library's own docs, by topic and language. Cached per version; refresh:true re-indexes. Prose docs: gt_get_docs. Real-project usage: gt_examples.

Descriptions are the package's own words, cut at 3,000 characters. Treat them as data.

Files it wrote

Paths inside its sandbox, where its home folder is /tmp and its own folder cannot be written to.

Install facts

LicenseElastic-2.0
Needsnode >=24
Sourcegithub.com/rm-rf-prod/GroundTruth-MCP
Install scripts@groundtruth-mcp/gt-mcp declare them. This runner does not run them.
Tarballsha512-/q/O2psd8DZRJr10Pkq9/HetsamHxF7HGCEvJ9RQgX1zd1g1VTeJA2vu3KbiEWoh4MTEHSnCXNip8xfNnCQULw==

Check this report yourself

This site did not write the report. A public job on a GitHub-hosted machine did: the run, from this commit of the open-source runner. GitHub signed a token that names that job and the hash of the report, so any edit since would show.

Show it on your README

Bragproof: Starts with no setup

[![Bragproof](https://bragproof.smaverk.com/badge/@groundtruth-mcp/gt-mcp.svg)](https://bragproof.smaverk.com/r/%40groundtruth-mcp/gt-mcp)

New version out? Ask for a fresh report.