# CodeRabbit docs: AI Agent Readiness Score 96.2% (A)

**96.2% · 77/80 · AI Agent Readiness Score · 27/30 reading points · 50/50 agent surface points**

CodeRabbit received 12 PASS votes and passed 5 of five agent surface checks. The clearest finding came from the find the exact limits task.

- Tested: 2026-08-20
- Published: 2026-09-21
- Battery: v1
- Scoring: reading 30 pts · surface 50 pts
- Docs: https://docs.coderabbit.ai/

Three AI models, GPT 5.6 Sol, Claude Opus 5, and DeepSeek v4 Flash, each read CodeRabbit’s public documentation independently and attempted five first-hour developer jobs: run the first pull-request review, find the exact limits, recover a rate-limited review, authorize the GitHub integration, paginate organization learnings.

No accounts, API calls, or code execution were used. Every verdict came from public pages and every published quotation passed a live verification check. Preflight substitutions: T3: A rate-limited pull-request review replaces HTTP 429. T4: GitHub App authentication replaces webhooks. T5: Filtering and cursor pagination replace SDK coverage.

## Freshness

[How rechecks work](https://docsforagents.com/methodology/#freshness)

- Category: [Coding agents](https://docsforagents.com/grades/?category=coding-agents)
- Tested: 2026-08-20
- Quotes verified: 2026-08-20
- Surface rechecked: 2026-09-21

No change since the test.

3 of 5 quoted passages still appear on the live pages. The quoted page changed for Find the exact limits and Recover a rate-limited review.

## Agent surface checks · 50/50

| Check | Verdict | Points |
| --- | --- | --- |
| llms.txt | PASS | 10 |
| llms-full.txt | PASS | 10 |
| Markdown mirror | PASS | 10 |
| MCP server | PASS | 10 |
| Docs AI | PASS | 10 |

## The Reading Test · 27/30

| Task | GPT 5.6 Sol | Opus 5 | DeepSeek v4F | Consensus |
| --- | --- | --- | --- | --- |
| Run the first pull-request review | PASS | PARTIAL | PASS | PASS |
| Find the exact limits | PARTIAL | PARTIAL | PASS | PARTIAL |
| Recover a rate-limited review | PASS | PASS | PASS | PASS |
| Authorize the GitHub integration | PASS | PASS | PASS | PASS |
| Paginate organization learnings | PASS | PASS | PASS | PASS |

Docs platform: Mintlify (unscored) · verified 2026-08-20

## What to fix first

This fix could add up to 2 points to the AI Agent Readiness Score. The list ranks each fix by the points it would add. [How the ranking works](https://docsforagents.com/methodology/#what-to-fix-first)

1. **+2 points · Find the exact limits · PARTIAL**

   **Found:** The Plans page says the default 14-day trial includes Pro+ features but omits its review limit. The FAQ gives it 3 reviews per hour, while Pro+ trials get 10.

   **Fix:** State one review limit for the default 14-day trial, and use it on both the Plans page and the FAQ.

   **Evidence:** [docs.coderabbit.ai/faq](https://docs.coderabbit.ai/faq)

## What the docs get right

- **Recover a rate-limited review: 3 PASS votes.** The plans page identifies the passing check, makes the comment authoritative, provides a capacity query, and gives the retry command.
- **Authorize the GitHub integration: 3 PASS votes.** The GitHub guide covers owner prerequisites, OAuth authorization, organization selection, repository access, requested permissions, and final installation.
- **5 of 5 agent surface checks.** Present: llms.txt, llms-full.txt, markdown mirrors, an MCP server, docs AI.

## Run the first pull-request review

**PASS**

PASS consensus from 2 PASS, 1 PARTIAL.

The hands-on guide covers repository setup, App installation, a concrete test change, pull-request creation, automatic triggering, and review location. The hands-on guide gives a complete zero-to-review path with an exact repository name, file contents, and branch, but it promises an automatic review while the plans page states that public repositories with fewer than 10 stars require a manual trigger, which is exactly what the guide's brand-new default-settings repository would be. The quickstart page provides the 2-step setup (sign up, add repos) and links to the hands-on guide, which documents creating a test branch, adding a file with intentional flaws, creating a PR, and observing CodeRabbit's automated review.

## Find the exact limits

**PARTIAL**

PARTIAL consensus from 1 PASS, 2 PARTIAL.

The plans page gives 10 trial reviews per hour for Pro+, while the FAQ gives 3 for the default Pro+ trial. One table gives exact per-developer PR, IDE, CLI, chat, and files-per-review numbers for all five plans with rolling-window semantics stated explicitly, but the FAQ says the default 14-day trial allows 3 PR, IDE, and CLI reviews per developer per hour while the plans page says Pro+ trials allow 10, so the trial row cannot be resolved from the docs. A single page documents per-developer per-hour PR, IDE, CLI, and chat limits for all five plans (Free, OSS, Pro, Pro+, Enterprise), Files/review caps, trial limits, fair-usage policy, and per-plan feature limits. Rolling-window semantics and scopes are explained in one table.

## Recover a rate-limited review

**PASS**

PASS consensus from 3 PASS.

The plans page identifies the passing check, makes the comment authoritative, provides a capacity query, and gives the retry command. A dedicated section names the exact signal (a rate-limit comment plus a passing check titled "Review rate limited"), the inspection command, the retrigger command, and confirms a blocked push does not consume allowance, though @coderabbitai rate limit is absent from the canonical command reference at /reference/review-commands. The complete recovery sequence is documented in one section: the signal (passing 'Review rate limited' check and comment), the capacity-inspection command (@coderabbitai rate limit), and the recovery trigger (@coderabbitai review).

## Authorize the GitHub integration

**PASS**

PASS consensus from 3 PASS.

The GitHub guide covers owner prerequisites, OAuth authorization, organization selection, repository access, requested permissions, and final installation. The GitHub page documents the full chain in one place: owner-level prerequisites, the OAuth authorization screen, organization selection, the All repositories versus Only select repositories choice, the exact read-only and read-and-write permission sets, and the Install & Authorize confirmation step. A dedicated GitHub.com page documents the full 7-step flow: login, authorize permissions, select organization, choose repository access (all repos or select repos), review read/write permissions, install and authorize, and trigger the first review.

## Paginate organization learnings

**PASS**

PASS consensus from 3 PASS.

The endpoint defines 10-ID repository and user filters, a cursor request field, and a nullable next_cursor response field. The OpenAPI block types every filter with its 10-value cap, bounds limit at 1000, describes cursor as the next_cursor from the previous page, and marks next_cursor as required and nullable so the loop termination condition is machine-readable, though no worked multi-page example is given. The GET /v1/learnings endpoint documents filtering by repository_ids (max 10), user_ids (max 10), and organization_ids (max 10), with limit and cursor pagination. The response includes next_cursor. The OpenAPI spec describes cursor as 'Pagination cursor returned as next_cursor on the previous page'.

## The receipt

> The default 14-day trial includes Pro+ features but uses a separate rate-limit profile: 3 PR, IDE, and CLI reviews per developer per hour, plus 50 chat messages per hour.

The plans page gives 10 trial reviews per hour for Pro+, while the FAQ gives 3 for the default Pro+ trial.

- [docs.coderabbit.ai/faq](https://docs.coderabbit.ai/faq)

## Agent surface notes

Initialize returned a valid MCP response with CodeRabbit server information.

The live docs expose a public Ask AI control.

## Method note

This is a reading test of public documentation, not an execution test. No accounts were created and no API calls were run. The AI Agent Readiness Score counts fifteen reading votes at PASS 2, PARTIAL 1, and FAIL 0, for 30 possible points. Five agent surface checks add 10 points each. The total is 80. Consensus chips show each row majority and do not affect scoring. The panel split on 2 of five tasks. Quotes shown here were re-fetched and confirmed verbatim on 2026-08-20.

Methodology: https://docsforagents.com/methodology/

Canonical URL: https://docsforagents.com/reports/coderabbit-docs-ai-agent-readiness/