Skip to main content
Glama

Small Print (MCP)

Small Print check

Listed on M8ven: https://m8ven.ai/mcp/gostanos-smallprint-action-12nii2

Small Print keeps a record of the text that MCP servers, agent skills and plugins give your AI agent, and tells you when a change is graded high or critical. It fetches the lists of the public registries, reads the new versions it can, and grades each change it finds by a written rule: https://smallprint.dev.

  • Get an email in the next daily brief after a change to something you use is graded high or critical. Sign up with just your email address at https://smallprint.dev/start, from the watch box on any entry's page, or by typing it when npx smallprint check asks. Pin the servers and skills you use; by default the daily brief shows changes graded high or critical and high or critical advisories, by email or Discord, and choosing everything in settings brings every change. No password, and no credit card required.

  • Check what your agents have installed, free and with no account: npx smallprint@1.6.0 check --no-upload prints what it found and sends nothing; without the flag it asks before sending names and versions to get grades. Or paste a config file at https://smallprint.dev/check.

  • Plans: New accounts get 30 days of Pro, no credit card required. What each plan includes, and the prices, are at https://smallprint.dev/pricing.

  • New to this? A step-by-step guide in plain words: https://smallprint.dev/guide

The grading rules are printed at https://smallprint.dev/how-we-grade.

What is in this repository

  • This Action fails a build when a repository's MCP server versions, skill files or instruction files differ from its lock.

  • npx smallprint gate asks the record before a session whether a change to the small print of each server here that has a registry name is on record since the lock, or a high or critical advisory covers the version here; exit codes for a shell hook. CLI 1.6.0.

  • mcp/: smallprint-mcp, an MCP server with four read-only tools over the public record (npx -y smallprint-mcp; registry name dev.smallprint/smallprint).

  • Dockerfile builds and runs that server over stdio, for registries that start a server to check it answers.

  • cli/: the source of the smallprint command, from which the npm package is built, with its tests. npx smallprint@1.6.0 check --no-upload prints what it found and sends nothing.

  • vscode/: the Small Print extension for VS Code and Cursor, the same files that go to the stores.

  • tools/verify-chain.mjs: checks the record chain from the public API: every link, every signature, and the entries hash of every day whose lines are kept; node tools/verify-chain.mjs. Rows written before 4 October 2026 link, and their entries cannot be recomputed.

Related MCP server: indexagentica-mcp

Small Print check, as a GitHub Action

Fails the build when an MCP server's version, a skill's files or an instruction file such as CLAUDE.md, AGENTS.md or an .mcp.json differs from a lock you committed.

- uses: gostanos/smallprint-action@v1.19

Write the lock from the repository's directory on a machine that has the intended configuration, with --project so it holds only what lives in the repository (the .mcp.json, the .claude/ and .cursor/ folders, CLAUDE.md, AGENTS.md and the rest), and commit it:

npx smallprint lock --project
git add smallprint.lock

To see it working, gostanos/smallprint-demo passes on its main branch, and its open pull request fails because it changes the filesystem server from version 2026.7.4 to 2026.7.10, and the lock check names that change. The check compares versions and file fingerprints; it does not read tool descriptions.

The check is local: it reads the repository's config and instruction files, compares them with the lock, and sends nothing anywhere. It exits 2 when something changed and prints what. When the change is yours, run npx smallprint lock --project again and commit. A lock written without --project also holds the machine's home-directory entries, which a CI runner does not have, so that check would fail on every run; the lock records which kind it is and the check honours it.

Inputs: lockfile (default smallprint.lock), version (the CLI version, default 1.6.0).

With a code-scanning upload

The step can write what changed as a SARIF log, and GitHub's upload-sarif action turns each line into a code-scanning alert on the pull request. Needs the --sarif flag, which is in smallprint 0.1.4 and later. The upload needs the security-events: write permission, and if: always() lets it run after the check has failed, so the build still fails when something changed.

    permissions:
      contents: read
      security-events: write
    steps:
      - uses: actions/checkout@v4
      - uses: gostanos/smallprint-action@v1.19
        with:
          sarif: smallprint.sarif
      - uses: github/codeql-action/upload-sarif@v3
        if: always()
        with:
          sarif_file: smallprint.sarif

As a pre-commit hook

With pre-commit, the same check runs before every commit, offline. Add this to .pre-commit-config.yaml in the repository that holds the lock:

repos:
  - repo: https://github.com/gostanos/smallprint-action
    rev: v1.19
    hooks:
      - id: smallprint-locked   # runs smallprint check --locked; the commit stops when anything differs from the lock
      - id: smallprint-lock     # only when you ask: pre-commit run smallprint-lock --hook-stage manual

The hooks run the pinned version of the command with npx, so the machine needs Node 20 or later.

Available Tools

4 tools
advisories_forAdvisories that name an entryA
Read-onlyIdempotent

Every public security advisory on record that names one entry, each attributed to the database or report that published it, with its severity criterion and the affected version range. Use it when deciding whether a specific version is inside a known advisory, or after lookup_entry reported advisories; it adds nothing for an entry with none. Small Print attributes and never judges: the severity is the source's or a printed CVSS band. Read-only: one HTTPS GET to smallprint.dev per call, no account, no key, nothing about the caller sent, and the server or skill asked about is never run or contacted. Rate limited to one entry per request; a 429 answer says to wait a minute. A name not in the catalog returns a plain error, not a guess.

Inputs: name (required), version. Returns: url, total, advisories, canonicalName

changed_since_approvalHas the small print changed since it was approved?A
Read-onlyIdempotent

Yes or no, before using a server or skill: has its small print changed since the version, content hash or date that was reviewed? The answer opens with UNCHANGED, CHANGED or UNKNOWN, then the releases that changed it since and their worst grade, then whether the review can stand. Use it on every run when an approval is on file, instead of re-reading the tools; use lookup_entry when nothing was approved yet and changes_since to read the diffs after a CHANGED answer. UNKNOWN means the approved version is not on record or its small print was never read, so nothing is compared; treat it as no answer, not as safe. Read-only: one HTTPS GET to smallprint.dev per call, no account, no key, nothing about the caller sent, and the server or skill asked about is never run or contacted. Rate limited to one entry per request; a 429 answer says to wait a minute. A name not in the catalog returns a plain error, not a guess.

Inputs: name (required), approved (required). Returns: url, status, advisories, worstGrade, +4 more

changes_sinceChanges to an entry's small printA
Read-onlyIdempotent

The releases of one entry whose tool descriptions, schemas or instructions changed, each with its diff, its grade and the rule that graded it (rules at https://smallprint.dev/how-we-grade). Use it to read what actually changed, after lookup_entry or changed_since_approval said something did; use changed_since_approval instead when the question is only whether anything changed since an approved version. Filters: since keeps releases published on or after a date; min_severity drops changes below a grade (default low, so plain version bumps and identical releases are never listed). Returns at most 12 releases in text; the structured result carries all of them. Read-only: one HTTPS GET to smallprint.dev per call, no account, no key, nothing about the caller sent, and the server or skill asked about is never run or contacted. Rate limited to one entry per request; a 429 answer says to wait a minute. A name not in the catalog returns a plain error, not a guess.

Inputs: name (required), since, min_severity. Returns: url, total, releases, canonicalName

lookup_entryLook up an entry on the Small Print recordA
Read-onlyIdempotent

What the record holds for one MCP server, skill or plugin: versions on record, the tools read from the pinned version, how many releases changed the small print and the worst grade, and the advisories that name it. Use it first, when nothing about the entry is known yet, or to confirm an entry exists before the other tools; use changes_since for the diffs and advisories_for for advisory detail. Not for private or unpublished servers, which have no page. Read-only: one HTTPS GET to smallprint.dev per call, no account, no key, nothing about the caller sent, and the server or skill asked about is never run or contacted. Rate limited to one entry per request; a 429 answer says to wait a minute. A name not in the catalog returns a plain error, not a guess.

Inputs: name (required). Returns: url, kind, toolsRead, advisories, +5 more

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.6/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool targets a distinct aspect of the change-tracking domain: lookup_entry for record overview, changed_since_approval for binary approval status, changes_since for detailed diffs, and advisories_for for advisory details. The descriptions explicitly cross-reference each other and clarify when to use one over another, leaving no meaningful ambiguity.

    Naming Consistency3/5

    All names use snake_case, but the grammatical patterns are mixed: changes_since (noun_preposition), lookup_entry (verb_noun), changed_since_approval (adjective_preposition_noun), and advisories_for (noun_preposition). The lack of a uniform verb_noun convention makes the set readable but inconsistently named.

    Tool Count5/5

    Four tools is well-scoped for a read-only change and advisory tracking service. Each tool covers a distinct lifecycle operation—lookup, approval check, diff inspection, and advisory query—with no redundant tools.

    Completeness4/5

    The set covers the core monitoring workflow: entry lookup, change since approval, detailed diffs, and advisories. A minor gap is the lack of a discovery/search tool to find entries by name or list recent changes across the catalog, though agents can typically work around this with known entry names.

    Maintenance

    ActivityActive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Read-only MCP tools for coding agents to audit deployment targets, detect fabricated code, review backlog, database schema, analytics, ML models, architecture docs, and decision lenses.
      11
      13 npm
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Read-only remote tools let agents search a curated directory of skills, harnesses, MCP servers, tools, protocols and APIs, then fetch full entries plus long-form guides, comparisons, stacks and downloadable Agent Skills. Needs no install, API key or authentication over Streamable HTTP.
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      A read-only MCP server that lets an agent check whether a public MCP server's tool definitions have changed, and whether any locally pinned server has a change waiting for review. It also returns the exact setup steps for protecting servers in Claude Code, Cursor, Codex, VS Code and others, without ever exposing third-party tool text.
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Publishes four flat, schema-light tools that let AI agents (including strict providers) search, describe, and invoke any of 139 reverse-engineering capabilities behind a persistent REA server. Enables file fingerprinting, inspection, and evidence-linked investigation of shipped software through plain-English requests, read-only by default.
      4
      2
      MIT