mcpField
S
Sigstore Rekor

Sigstore Rekor

Remote
pipeworx.ioSecurity0Updated 5w ago

About Sigstore Rekor

Sigstore Rekor exposes the public transparency log for software signing records, for developers and security teams verifying provenance. Its tools retrieve log state and look up entries by index, artifact hash, email, or UUID.

Key features

  1. 01Inspect log state and Merkle root
  2. 02Fetch entries by index or UUID
  3. 03Search by artifact hash or email
  4. 04Decode entry bodies and inclusion proofs

Use cases

  1. 01Check whether an artifact has a public signing record
  2. 02Review a signing certificate and inclusion proof
  3. 03Inspect Rekor log state before querying entries

MCP support

Provides MCP tools for querying Sigstore Rekor through the Pipeworx gateway or as a standalone stdio server.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://gateway.pipeworx.io/sigstore-rekor/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

@pipeworx/sigstore-rekor

The Sigstore Rekor public transparency log — the immutable, append-only record of software signing events behind cosign, npm and PyPI provenance, and SLSA attestations.

Part of Pipeworx — an MCP gateway connecting AI agents to 1683+ live data sources.

Tools

  • rekor_log_info(...) — current log state: total entries, signed tree head, Merkle root, retired shards. Call it for the upper bound on log_index.
  • rekor_entry_by_index(log_index) — one entry by its integer position, body decoded.
  • rekor_search_by_hash(hash | email) — given a sha256 artifact digest (or a signing-certificate email identity), every entry that signed it. Returns uuids.
  • rekor_entry(uuid) — one entry by uuid, with signing certificate, integration time and inclusion proof.

Auth

Keyless.

Data sources

Notes the next person would otherwise rediscover:

  • body is base64-encoded JSON. Every tool here decodes it into entry_body and lifts the artifact digest and entry kind to the top level; a caller that reads the raw field gets an opaque blob.
  • logIndex and uuid are different identifier spaces. rekor_search_by_hash returns uuids, so it feeds rekor_entry, not rekor_entry_by_index.
  • Entry KINDS (rekord, hashedrekord, intoto, dsse) do not share a body shape. summarize() looks each field up where that kind puts it and returns null where the kind has no equivalent — do not assume artifact_digest is always populated.
  • The index search hashes want the sha256:<hex> form. A bare hex digest is what every other tool prints, so both are accepted and normalised in-pack.
  • No entry for a digest is NOT evidence the artifact is bad — only that there is no public signing record. The tool says so in its note.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "sigstore-rekor": {
      "url": "https://gateway.pipeworx.io/sigstore-rekor/mcp"
    }
  }
}
What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/sigstore-rekor/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1683+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/rekor_log_info \
  -H 'Content-Type: application/json' \
  -d '{}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/rekor_log_info. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "sigstore-rekor": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-sigstore-rekor"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-sigstore-rekor

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Sigstore Rekor data" })

The gateway picks the right tool and fills the arguments automatically.

More

License

MIT

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.