Pydantic AI by Pydantic

# Fix list: Pydantic AI

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/pydantic-ai, the October 2026 research run, assessed 1 October 2026. Grade A, 80 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 Pydantic 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. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: No payment protocol (0). Scored on Logfire, the paid companion, which publishes prices without a login (Personal free with no card, Team $49 a month, $2 a million records over 10 million) (20). The MIT package installs with no card (20) and no account, and the test model needs no key at all (20).

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.

## 2. Security & auth, 80 out of 100, up to 3.5 more on the total

Why it scored 80: No telemetry unless you configure it. OpenTelemetry instrumentation and Logfire take two added lines (30). Human approval is built in, but we found no read-only mode or sandbox for model-written code (10). Guardrails come in the Harness, and output validation retries, but we found no prompt-injection guidance (10). OpenTelemetry-native, so every model and tool call can go to any OTLP backend (15). SECURITY.md uses GitHub private reporting, no bounty is mentioned, and seven advisories were published in 2026 with fixed versions (15). Framework reading, so SOC 2 isn't scored.

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. Reliability, 83 out of 100, up to 3.4 more on the total

Why it scored 83: Official package on PyPI (Requires-Python >=3.10) (20). CI passes on main, with a coverage badge (25). 560 open issues and 219 open pull requests (8). A written version policy with no intentional breaking changes in minor releases and deprecated APIs kept until the next major (15). 2.52.0, classed 5 - Production/Stable (15).

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.

## 4. Agent ergonomics, 85 out of 100, up to 2.4 more on the total

Why it scored 85: MCP ships in core, but we didn't see tool filtering or the length of the minimal example this run (15). Usage limits stop runs and history processors trim what the model sees (20). Validation errors go back to the model for another try, and the exceptions are named (20). Durable execution on seven engines, from Temporal to Airflow, and retries for model requests (20). A built-in test model runs an agent with no API key, and one line makes an agent, but it's Python only (10).

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

Made of editorial 90, provenance 63.

Why it scored 77: MIT (30). The overview says instrumentation is opt-in and works with any OTLP backend, and Logfire's plans state their limits, but we didn't find a page for the library that says in so many words what leaves the machine (20). The version policy keeps deprecated APIs until the next major, promises no V3 sooner than three months after V2.0 and V1 security fixes for at least six months after V2 on 2026-06-23 (20). Telemetry is opt-in and documented (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: pydantic.dev, registered 2022-04-24 (4 years) (7 of 15)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

## 6. Maintenance & community, 90 out of 100, up to 0.9 more on the total

Why it scored 90: 2.52.0 on 2026-09-30 (30). More than 50 releases since 2026-07-03 (20). 560 open issues and 219 open pull requests, and we couldn't see reply times (15). The Python package is current (15). CI and coverage pass on main (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. Schema & documentation, 95 out of 100, up to 0.8 more on the total

Why it scored 95: Typed end to end with an API reference (25). llms.txt, per the listing's earlier check (10). The docs separate agents, graphs and the Harness, though we didn't re-check the when-not-to-use wording this run (15). Tools are typed functions validated by Pydantic (15). `ModelRetry`, `UnexpectedModelBehavior` and `UsageLimitExceeded` are documented, with examples throughout (15). Changelog and a version policy (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.

## Deductions

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

- 2026-02-06. Two high-severity advisories, server-side request forgery in URL download handling (GHSA-2jrp-274c-jhv3, CVE-2026-25580) and stored XSS through path traversal in the web UI's CDN URL (GHSA-wjp5-868j-wqv7). Both fixed and published, and almost eight months old, so 1 point each. Five moderate advisories from May to August 2026, two of them bypasses of its cloud-metadata blocklist, weren't deducted. https://github.com/pydantic/pydantic-ai/security

## 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.

- We couldn't load pydantic.dev's terms and privacy pages this run, so the provenance block's blank fields are unchanged
- We didn't confirm the MCP page's tool filtering or example length this run
- PyPI's release list gave between 52 and 58 releases since 2026-07-03 in two readings, so we wrote more than 50
- We didn't find a page that states, for the library, that nothing is sent without configuration, only that instrumentation is opt-in

## Weaknesses

- Python only
- 560 open issues and 219 open pull requests
- Seven advisories in 2026, including SSRF and two bypasses of its cloud-metadata blocklist
- No sandbox for model-written code
- Guardrails live in the separate Harness

## 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.

- Define the output type first. Validation retries fix most malformed answers without a prompt change
- Start with the test model to check the wiring without a key
- Use streamable HTTP for MCP. SSE is deprecated
- Set usage limits on every run that calls paid models
- Stay on a current release if agents download URLs. Its cloud-metadata blocklist was bypassed twice in May 2026

## What the review panel asked for

- Document data egress
- MCP tool filtering documented
- Sandbox for generated code
- State the usage limit unit
- page on outbound data
- Document retry counts and timeout defaults in one page
- sandbox for generated code
- prompt-injection guidance
- a dated V3 announcement
- Show tool filtering on the MCP page

## 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.