Puppeteer (archived MCP reference server) by Model Context Protocol (archived)

# Fix list: Puppeteer (archived MCP reference server)

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/puppeteer-reference-server-archived, the October 2026 research run, assessed 1 October 2026. Grade F, 30.8 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Puppeteer (archived MCP reference server): work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Reliability, 12 out of 100, up to 17.6 more on the total

Why it scored 12: Archived and superseded, so we scored what's left as a local stdio package. @modelcontextprotocol/server-puppeteer 2025.5.12 still installs from npm, deprecated there with a generic message, and no runtime is stated (10). No tests and no CI for it in the archive (0). The repository has been read-only since 29 May 2025, so nobody can file or fix a crash, and console logs collect in an array that never empties (0). A date version and no changelog (2). Never 1.0 and never declared stable (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 2. Security & auth, 22 out of 100, up to 13.7 more on the total

Why it scored 22: No credentials to leak and nothing to scope, so the middle band (20). The guard against dangerous Chrome flags such as `--no-sandbox` is switched off by `allowDangerous: true` in the `puppeteer_navigate` arguments, so the model or a page that steers it can lift it. The Docker mode always launches with `--no-sandbox --single-process --no-zygote`, and `puppeteer_evaluate` runs any script (2). Page content and console output reach the model unmarked, with no injection guidance. The README's only caution is that the browser can reach local files and internal addresses (0). No call log (0). The archive README says "NO SECURITY GUARANTEES", and the pinned Puppeteer ^23.4.0 is itself deprecated on npm as "< 24.15.0 is no longer supported" (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 3. Schema & documentation, 41 out of 100, up to 9.6 more on the total

Why it scored 41: Seven tools with JSON Schema inputs and `required` lists on every one, but `launchOptions` is a free-form object (20). The README is Markdown on GitHub, no llms.txt (5). Descriptions are three to nine words, "Navigate to a URL", "Execute JavaScript in the browser console", with no when-to-use guidance (4). No enums or constraints. Selectors and scripts are free strings (6). The README shows `launchOptions` examples, and failures come back as `isError` text such as "Script execution failed" (6). No versioning or changelog beyond the date version (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 4. Maintenance & community, 0 out of 100, up to 8.8 more on the total

Why it scored 0: Last release 2025.5.12 on 12 May 2025 (0). No releases in the last 90 days (0). Archived on 29 May 2025 and read-only (0). Not in the official MCP registry (0). Pinned to MCP SDK 1.0.1 and an unsupported Puppeteer major (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## 5. Agent ergonomics, 47 out of 100, up to 8.6 more on the total

Why it scored 47: Seven compact tools, about 700 tokens in total (25). No snapshot or text extraction, so the agent works from screenshots and `puppeteer_evaluate`. Screenshot `width` and `height` are the only size controls (4). Errors return as tool results with the browser's message (10). No annotations at all (0). One or two required parameters per tool, Node only (8).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 6. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: Free, self-hosted, nothing to buy, so 20 + 20 + 20. No payment protocol (0). The rubric is mechanical here and says nothing in favour of installing it.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 7. Transparency & trust, 77 out of 100, up to 2 more on the total

Made of editorial 80, provenance 74.

Why it scored 77: MIT, an OSI licence (30). Local software that keeps screenshots and console logs in memory and talks only to the pages it opens (20). The GitHub banner dates the archive to 29 May 2025 and the archive README says no security updates will follow. Neither the archive, the main repository's Archived list nor the npm deprecation message names a successor, where the Brave and Slack entries do (10). No telemetry (20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: modelcontextprotocol.io, registered 2024-11-18 (1 year) (3 of 15)
- Status page: not found (0 of 10)

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- -4: an open exfiltration path that will never be fixed. The model can set `allowDangerous: true` on `puppeteer_navigate` to relaunch Chrome with the sandbox off, `puppeteer_evaluate` runs any script, and the README warns the browser can read local files and internal addresses. The package still drew 25,072 npm downloads in the week of 14 to 20 August 2026, the latest week the npm API returned to us. Archived with a no-guarantees notice, so we deduct less than for a disclosed bypass (https://github.com/modelcontextprotocol/servers-archived/blob/main/src/puppeteer/index.ts; https://api.npmjs.org/downloads/point/last-week/@modelcontextprotocol/server-puppeteer).

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- The npm downloads API returned the week of 14 to 20 August 2026 for a last-week query on 1 October, so the current weekly figure may differ
- unchecked: whether the mcp/puppeteer Docker image is still published on Docker Hub
- unchecked: the date npm deprecated the package, since the registry record carries the message but no date

## Weaknesses

- Archived on 29 May 2025 with no security guarantees, and deprecated on npm
- `allowDangerous: true` in a tool call lifts the sandbox guard, and Docker mode never had one
- Pins Puppeteer ^23.4.0, which npm marks as no longer supported
- No accessibility snapshot or text extraction, so the agent works from screenshots and scripts
- No tool annotations, no tests, no successor named anywhere official

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Don't install it for new work. playwright-mcp covers the same tasks with snapshots and annotations
- If you inherit a config for it, set ALLOW_DANGEROUS to false and never pass `allowDangerous`
- Never run it with a logged-in profile or on a host with internal services the browser can reach

## What the review panel asked for

- Name a successor
- named successor notice

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.