# LlamaIndex docs: AI Agent Readiness Score 71.2% (C-)

**71.2% · 57/80 · AI Agent Readiness Score · 27/30 reading points · 30/50 agent surface points**

LlamaIndex received 12 PASS votes and passed 3 of five agent surface checks. The clearest finding came from the use the Python SDK task.

- Tested: 2026-09-22
- Published: 2026-09-22
- Battery: v1
- Scoring: reading 30 pts · surface 50 pts
- Docs: https://developers.llamaindex.ai/python/framework/

Three AI models, GPT 5.6 Sol, Claude Opus 5, and DeepSeek v4 Flash, each read LlamaIndex’s public documentation independently and attempted five first-hour developer jobs: index and query a document, find the exact limits, recover from a 429, verify a webhook, use the Python SDK.

No accounts, API calls, or code execution were used. Every verdict came from public pages and every published quotation passed a live verification check.

## Freshness

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

- Category: [Agent frameworks & memory](https://docsforagents.com/grades/?category=agent-frameworks)
- Tested: 2026-09-22
- Quotes verified: 2026-09-22
- Surface rechecked: Not yet rechecked

## Agent surface checks · 30/50

| Check | Verdict | Points |
| --- | --- | --- |
| llms.txt | PASS | 10 |
| llms-full.txt | FAIL | 0 |
| Markdown mirror | PASS | 10 |
| MCP server | PASS | 10 |
| Docs AI | FAIL | 0 |

## The Reading Test · 27/30

| Task | GPT 5.6 Sol | Opus 5 | DeepSeek v4F | Consensus |
| --- | --- | --- | --- | --- |
| Index and query a document | PASS | PASS | PASS | PASS |
| Find the exact limits | PARTIAL | PASS | PASS | PASS |
| Recover from a 429 | PASS | PASS | PASS | PASS |
| Verify a webhook | PASS | PASS | PASS | PASS |
| Use the Python SDK | PARTIAL | PARTIAL | PASS | PARTIAL |

Docs platform: Astro Starlight (unscored) · verified 2026-09-22

## What to fix first

These 3 fixes could add up to 22 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. **+10 points · llms-full.txt · check failed**

   **Found:** HTTP 404 returning the Starlight 404 HTML page.

   **Fix:** Publish llms-full.txt at the docs root with the full text of every docs page in one plain-text file.

   **Evidence:** [developers.llamaindex.ai/llms-full.txt](https://developers.llamaindex.ai/llms-full.txt)

2. **+10 points · Docs AI · check failed**

   **Found:** The docs load a RunLLM widget script, but the #runllm-widget container stayed empty with no openable control after 47 seconds.

   **Fix:** Add an assistant to the docs site that answers questions from the docs and links to its sources.

3. **+2 points · Use the Python SDK · PARTIAL**

   **Found:** The Python SDK reference documents only parsing.create, while the platform quickstart and the Parse examples call client.parsing.parse, a method the reference never lists.

   **Fix:** List client.parsing.parse in the Python SDK reference or change the quickstart examples to parsing.create, and show the source_url alternative to file_id.

   **Evidence:** [developers.llamaindex.ai/reference/python/resources/parsing/methods/create](https://developers.llamaindex.ai/reference/python/resources/parsing/methods/create/)

## What the docs get right

- **Index and query a document: 3 PASS votes.** The framework landing page gives the API key variable, the install command, and a five-line starter in one sequence.
- **Recover from a 429: 3 PASS votes.** The troubleshooting page names the 429, states the free-tier limit, and prescribes exponential backoff with jitter.
- **3 of 5 agent surface checks.** Present: llms.txt, markdown mirrors, an MCP server. Missing: llms-full.txt, docs AI.

## Index and query a document

**PASS**

PASS consensus from 3 PASS.

The framework landing page gives the API-key variable, pip command, local data folder, indexing call, query engine, and query call in one sequence. One linear path runs from pip install llama-index through export OPENAI_API_KEY=XXXXX, a wget that creates the sample data/ directory, and a complete copy-pasteable starter.py that indexes and queries it. The homepage quickstart section documents a single clear path from install to query in five lines.

## Find the exact limits

**PASS**

PASS consensus from 2 PASS, 1 PARTIAL.

The docs state parsing prices and monthly credits, but the rate table covers only the v1 parse-upload route while the current guide uses v2. The Rate Limits page names the Parse upload route, its numeric limit, and its scope, and the free-tier figure of 20 requests per minute matches the plan table on the Billing page, while Pricing gives parse credit costs per tier against a stated $1.25 per 1,000 credits. Rate limits, credit costs per tier, and plan quotas are documented and consistent across the rate limits, pricing, and billing pages.

## Recover from a 429

**PASS**

PASS consensus from 3 PASS.

The troubleshooting page names 429, directs readers to endpoint limits, prescribes jittered exponential backoff, suggests batching, and includes a status-code reference. A dedicated Troubleshooting & Error Codes page carries a 429 section, per-product error-message tables mapping exact strings to status codes and resolutions, and an HTTP status code reference, and it states the absence of a Retry-After header rather than leaving the reader to discover it. The troubleshooting page provides a dedicated 429 section referencing the rate limits page, plus an error reference table with explicit 429 messages and resolution guidance.

## Verify a webhook

**PASS**

PASS consensus from 3 PASS.

One page documents endpoint configuration, completion events, payload and headers, raw-body HMAC-SHA256, and a Python receiver. The page covers configuration inline and by saved config ID, the full job-completion payload and header list, and runnable Python and TypeScript verifiers that use a constant-time comparison and warn against re-serializing the parsed JSON before hashing. Webhook configuration, payload structure, event types, signature verification code, and a complete FastAPI receiver example are all documented on one page.

## Use the Python SDK

**PARTIAL**

PARTIAL consensus from 1 PASS, 2 PARTIAL.

The platform quickstart is end to end, but the generated parsing.create() example omits both source alternatives required by its own reference text. The SDK is easy to find and its quickstart runs, but the reference documents only parsing.create while the platform quickstart and every Parse example call client.parsing.parse(...), a method the Python SDK reference never lists, so the example form and the reference form disagree. The SDK reference page provides a minimal working example that matches the client and method form used across the parse getting-started pages and LlamaCloud platform documentation.

## The receipt

> Provide either file_id (a previously uploaded file) or source_url (a publicly accessible URL).

The generated reference states the two input options but leaves the upload step to a separate page.

- [developers.llamaindex.ai/reference/python/resources/parsing/methods/create/](https://developers.llamaindex.ai/reference/python/resources/parsing/methods/create/)

## Agent surface notes

Initialize returned JSON-RPC protocol 2025-06-18 and serverInfo for the LlamaIndex documentation server.

The docs load a RunLLM widget script, but the #runllm-widget container stayed empty with no openable control after 47 seconds.

## 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 2026-09-22.

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

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