mcpField
G
guardfile

guardfile

Local
github.com/CloudCorpRecordsVersion Control0Updated 1w ago

About guardfile

guardfile checks supplied repository files for four contribution guardrails: DCO, CLA, changesets, and sign-off. It returns a report with evidence paths, missing checks, and an overall completeness flag, using only the provided content.

Key features

  1. 01Detect DCO, CLA, changesets, and sign-off
  2. 02Match guardrails by file path or content
  3. 03List sorted evidence paths without duplicates
  4. 04Report present and missing checks
  5. 05Validate input locally

Use cases

  1. 01Check contribution requirements before a release
  2. 02Review repository files for guardrail evidence
  3. 03Identify missing DCO, CLA, changeset, or sign-off rules

MCP support

The MCP server exposes an `extract_guardrails` tool that accepts repository-relative paths and file contents, then returns detected guardrails, evidence, missing items, and completeness.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"cloudcorprecords-guardfile": {
"command": "npx",
"args": [
"-y",
"@reneturcios/guardfile"
]
}
}
}
Run commandnpx -y @reneturcios/guardfile

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@reneturcios/guardfile0.1.0
ocighcr.io/cloudcorprecords/guardfile:0.1.0

Tools

1

Saved tool list · updated 1w ago

ParameterTypeRequiredDescription
filesarrayYesRepo files to inspect (at least one)

Details

From the project README

guardfile

A repo can claim all sorts of contribution guardrails — sign your commits, agree to the licence terms, add a changeset — but the claim only means something if you can point at the files that put each one in force. guardfile does that pointing: you hand it a repo's files, and it hands back one checkable report of which guardrails are actually present, with the evidence for each.

Four guardrails are detected, by fixed rules:

  • DCO — a file whose basename is exactly DCO (any case), or any content naming the Developer Certificate of Origin.
  • CLA — a file whose basename starts with CLA (any case, as in CLA.md), or any content naming the Contributor License Agreement.
  • changesets — a path containing .changeset/, a path ending in /.changeset, a path that is exactly .changeset, or any content naming the @changesets/ scope.
  • sign-off — any content carrying a Signed-off-by: line. The colon is part of the rule: a bare Signed-off-by mention does not count.

All content matching is case-insensitive. For each guardrail the report lists the paths that triggered it, sorted, with each path appearing at most once even when a single file triggers the same guardrail by both its path and its content. Alongside the four verdicts the report gives the number present, the names of the ones missing (in the fixed order dco, cla, changeset, signoff), and a complete flag that is true only when all four are present.

Malformed input — a missing or empty file list, an entry that is not an object, a missing, empty, or duplicated path, or content that is missing or not a string — is a validation error naming the offending path (for example files[1].content must be a string), and no partial report is produced. The computation is pure and local: no network, no configuration, and no files read beyond the input you pass in.

CLI

npx @reneturcios/guardfile extract --input args.json
npx @reneturcios/guardfile --version

args.json contains the full arguments object:

{
  "files": [
    { "path": "DCO", "content": "Developer Certificate of Origin\n\nVersion 1.1" },
    { "path": "docs/contributing.md", "content": "Please add a Signed-off-by: line." },
    { "path": ".changeset/bright-foxes-jump.md", "content": "---\n\"pkg\": patch\n---\n" }
  ]
}

The report is printed as JSON:

{
  "guardrails": {
    "dco": { "present": true, "evidence": ["DCO"] },
    "cla": { "present": false, "evidence": [] },
    "changeset": { "present": true, "evidence": [".changeset/bright-foxes-jump.md"] },
    "signoff": { "present": true, "evidence": ["docs/contributing.md"] }
  },
  "present_count": 3,
  "missing": ["cla"],
  "complete": false
}

Exit codes: 0 on success, 1 on a runtime error (including an unreadable or invalid input file, or invalid input), 2 on a usage error.

MCP server

npx -y @reneturcios/guardfile mcp

MCP client config:

{
  "mcpServers": {
    "guardfile": {
      "command": "npx",
      "args": ["-y", "@reneturcios/guardfile", "mcp"]
    }
  }
}
Tools
  • extract_guardrails — Extract a repo's guardrails from its files.

    • files (array, required, at least one) — repo files as {path, content} objects; path is a repo-relative path unique across the set and content is the file's content, which may be empty

    Returns {guardrails, present_count, missing, complete}.

Docker

docker run -i ghcr.io/cloudcorprecords/guardfile
docker run --rm -v "$PWD:/w" ghcr.io/cloudcorprecords/guardfile extract --input /w/args.json

Development

npm install
npm run build   # tsc -> dist/
npm test        # tsx --test test/*.test.ts (pure local, no network)
npm run dev     # run the CLI from source

License

MIT — Copyright (c) 2026 Rene Turcios

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.