Guardrails AI by Guardrails AI (Harvey)

# Fix list: Guardrails AI

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/guardrails-ai, the October 2026 research run, assessed 1 October 2026. Grade D, 49.8 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 Guardrails AI: 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. Security & auth, 44 out of 100, up to 9.8 more on the total

Why it scored 44: Framework reading. No auth of its own, and the server has none built in. Provider keys come from the environment (10 of 30). Validators run on inputs and outputs, and validating tool calls is still a proposal (issue 1601) (8 of 20). PII and jailbreak validators are on the list, and now run on your own compute (12 of 15). Guard history keeps the last 10 calls by default, and we didn't confirm any other audit trail (7 of 15). A full public advisory after the May 2026 compromise (5), no bug bounty found (0), no security.txt or disclosure policy checked (0), and metrics on by default in the client config with the disclosure not confirmed (2), so 7 of 20.

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.

## 2. Reliability, 58 out of 100, up to 8.4 more on the total

Why it scored 58: Local framework reading. Installs from PyPI as guardrails-ai, Python 3.10 to 3.13 stated (20). A CI workflow badge on the README, but we didn't confirm the default branch passes (15 of 25). 38 open issues, the newest from 25 July 2026, including bug reports, and we couldn't see reply rates (15 of 25). GitHub releases mark breaking changes (0.8.0, 0.8.1, 0.9.0), but the newest release on the releases page is 0.10.2 while PyPI has 0.11.0 from 14 August 2026 with no release notes we could find (8 of 15). Version 0.11.0, and the open 1.0.0 issues plan to delete reask, on_fail, RAIL and structured decoding (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.

## 3. Schema & documentation, 59 out of 100, up to 6.7 more on the total

Why it scored 59: Framework reading. Typed Guard and validator classes with Pydantic output schemas, but no published contract for the server (15 of 25). No llms.txt found (0). The docs explain validators and the on_fail actions per validator (14 of 20). Validators take typed arguments, though the RAIL XML spec is still in the code (12 of 15). Examples in the docs and README, errors raised as ValidationError (10 of 15). Semver tags, but 0.11.0 has no entry on the releases page (8 of 15).

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. Agent ergonomics, 63 out of 100, up to 6 more on the total

Why it scored 63: Framework reading. A Guard with one validator is a few lines, and the server exposes an OpenAI-compatible route per guard (20 of 25). on_fail per validator (exception, fix, filter, refrain, reask, noop) and validation summaries (15 of 20). Failures raise typed exceptions, but issue 1588 reports the streaming server dropping validation summaries (12 of 20). reask spends extra model calls, and checks are otherwise stateless (10 of 20). Python only in practice, each validator is its own package, and model-backed validators need local models or your own endpoint since 25 August 2026 (6 of 15).

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.

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

Why it scored 60: Apache-2.0 package you run yourself. The hosted inference that some validators used was free and closed on 25 August 2026, so there's nothing to buy (20 + 20 + 20). No payment protocol (0).

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.

## 6. Maintenance & community, 44 out of 100, up to 4.9 more on the total

Why it scored 44: 0.11.0 on PyPI on 14 August 2026, 48 days ago (20). One release in the last 90 days (0). The newest open issue is from 25 July 2026 and we couldn't see maintainer replies, and the company was bought by Harvey on 9 September with no word on the library (8 of 25). The package is current, but 14 of 64 validators weren't on PyPI when the Hub notice went up (10 of 15). CI exists, Python 3.10 to 3.13, but use_remote_inferencing still defaults to true in the client config while the hosted endpoints are gone (6 of 10).

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.

## 7. Transparency & trust, 61 out of 100, up to 3.4 more on the total

Made of editorial 63, provenance 59.

Why it scored 61: Apache-2.0 (30). With the Hub closed nothing passes through Guardrails AI servers except metrics, but we couldn't find what the metrics contain, and the privacy policy (updated 14 August 2025) predates the Harvey acquisition (12 of 30). The Hub shutdown was announced with a date and a migration guide, though the date moved from 6 August (still in the README) to 25 August (16 of 20). enable_metrics defaults to true in ~/.guardrailsrc and can be set false, but we didn't find this in the docs (5 of 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: guardrailsai.com, no registry record we could read (0 of 15)
- Status page: not found (0 of 10)
- security.txt: could not be fetched (0 of 10)

## Deductions

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

- 2026-05-11 supply-chain compromise. An attacker used an employee's GitHub token to run Actions across 30 repositories, took deploy secrets and published a malicious guardrails-ai 0.10.1 to PyPI. Quarantined in about two hours, tokens rotated, Hub and Snowglobe keys force-rotated on 13 May, and a full advisory published telling anyone who installed 0.10.1 to treat the host as compromised. Fixed and documented, so partly decayed (-6). https://github.com/guardrails-ai/guardrails/blob/main/SECURITY_ADVISORY.md

## 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.

- What Harvey plans for the open-source library and the guardrailsai.com docs.
- What the default-on metrics collect, and whether they still reach a Guardrails AI endpoint after the Hub closed.
- Whether the default branch CI passes, and whether maintainers still answer issues. The newest open issue is from 25 July 2026.
- Whether 1.0.0 will ship, given the issues filed on 24 July 2026.

## Weaknesses

- Acquired by Harvey on 9 September 2026 with no statement on the library
- Hub, private registry and hosted inference closed on 25 August 2026, so model-backed validators need your own compute
- Malicious 0.10.1 release on PyPI in May 2026 from a compromised token
- 0.11.0 has no release notes on GitHub, and open 1.0.0 issues plan to remove reask, on_fail and RAIL
- Metrics on by default in the client config

## 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.

- Pin guardrails-ai==0.11.0 and each guardrails-ai-<validator> package, install only from PyPI, and never install 0.10.1
- Import validators from guardrails_ai.<name>, not guardrails.hub, and don't run guardrails hub install
- Pass use_local=True to detect_pii, toxic_language and the other model-backed validators, or set validation_endpoint to a server you run
- Set enable_metrics to false in ~/.guardrailsrc if you don't want usage metrics sent
- Avoid building on reask and RAIL. The open 1.0.0 issues plan to remove both

## What the review panel asked for

- Correct the README Hub date
- Add release notes for 0.11.0
- a documented metrics payload
- a published disclosure policy

## 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.