Published 2026-09-22 · Tested 2026-09-22
Doppler
BDoppler received 12 PASS votes and passed 4 of five agent surface checks. The clearest finding came from the recover from a 429 task.
Three AI models, GPT 5.6 Sol, Claude Opus 5, and DeepSeek v4 Flash, each read Doppler’s public documentation independently and attempted five first-hour developer jobs: inject the first secret, find the exact limits, recover from a 429, verify a webhook, use the Node 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- Category
- Security & compliance
- Tested
- Quotes verified
- Surface rechecked
- Not yet rechecked
83.8% · 67/80 · AI Agent Readiness Score · reading 30 pts · surface 50 pts
Agent surface checks · 40/50
The Reading Test
| Task | GPT 5.6 Sol | Opus 5 | DeepSeek v4F | Consensus |
|---|---|---|---|---|
| Inject the first secret | PASS | PASS | PASS | PASS |
| Find the exact limits | PARTIAL | PASS | PASS | PASS |
| Recover from a 429 | PARTIAL | PARTIAL | PASS | PARTIAL |
| Verify a webhook | PASS | PASS | PASS | PASS |
| Use the Node SDK | PASS | PASS | PASS | PASS |
docs platform: ReadMe (unscored) · verified 2026-09-22
What to fix first
These 2 fixes could add up to 12 points to the AI Agent Readiness Score. The list ranks each fix by the points it would add. How the ranking works
- 1 +10 points llms-full.txt check failed
Found: HTTP 404 returning the ReadMe HTML shell at both the documentation root and the /docs path.
Fix: Publish llms-full.txt at the docs root with the full text of every docs page in one plain-text file.
Evidence: docs.doppler.com/
llms-full.txt (opens in a new tab) - 2
Found: The API reference documents retry-after and the 429 response, but no header or page says whether the IP limit or the token limit fired.
Fix: In the API reference, state which requests the IP limit covers and which the token limit covers, and map the 240 x-ratelimit-limit value to each plan and operation.
Evidence: docs.doppler.com/
reference/ api (opens in a new tab)
What the docs get right
- Inject the first secret: 3 PASS votes. The CLI guides document secret creation, login, project/config selection, and process injection through a short command sequence.
- Verify a webhook: 3 PASS votes. One guide covers dashboard creation, the default event payload, signature generation, and timing-safe verification examples.
- 4 of 5 agent surface checks. Present: llms.txt, markdown mirrors, an MCP server, docs AI. Missing: llms-full.txt.
Inject the first secret
PASSPASS consensus from 3 PASS.
The CLI guides document secret creation, login, project/config selection, and process injection through a short command sequence. One page carries the whole path: per-OS install, doppler --version, doppler login, doppler setup, then doppler run -- your-command-here. The install-cli page documents the complete linear path: install, verify, doppler login, doppler setup, doppler run -- your-command-here.
Find the exact limits
PASSPASS consensus from 2 PASS, 1 PARTIAL.
Both tables match, but the API also names an IP-based limit without giving its exact value. Platform Limits and the API Reference state the same exact per-plan numbers (reads 240/480/480 per min, secret reads 120/240/480, writes 60/120/240), and Platform Limits adds exact plan, abuse, and name-length ceilings. The API reference and the platform-limits page both publish the identical per-plan rate limit numbers, with no contradiction between them.
Recover from a 429
PARTIALPARTIAL consensus from 1 PASS, 2 PARTIAL.
The headers provide a limit number and retry timing, but they do not identify the triggering scope or request bucket. retry-after and x-ratelimit-reset give a wait, but no header distinguishes the IP-address limit from the token limit and an x-ratelimit-limit of 240 maps to three different plan-and-operation combinations, so identifying which limit was hit requires guessing. The API reference describes two rate limit types (IP- and key-based), documents all four rate-limit headers including retry-after, and states that HTTP 429 signals a hit.
Verify a webhook
PASSPASS consensus from 3 PASS.
One guide covers dashboard creation, the default event payload, signature generation, and timing-safe verification examples. Creation steps, the full config.secrets.update payload example, the sha256= prefix convention, and timing-safe Node and Ruby verification code all sit on one page, with the API's webhook secret field linking back to that same anchor. One page covers creating a webhook, the config.secrets.update default payload, the X-Doppler-Signature header, and Node.js/Ruby verification code.
Use the Node SDK
PASSPASS consensus from 3 PASS.
The SDK's projects.list() call matches the reference's GET https://api.doppler.com/v3/projects operation. The example's doppler.projects.list() corresponds to GET /v3/projects with page and per_page in the API reference, so the SDK call and the REST form agree. The GitHub README example calls sdk.projects.list(), which the SDK reference maps to GET /v3/projects, matching the REST API reference endpoint.
The receipt
Suggested wait period in seconds. This is only returned after a rate limit is hit.
The headers provide a limit number and retry timing, but they do not identify the triggering scope or request bucket.
Agent surface notes
Initialize returned HTTP 401 carrying a JSON-RPC 2.0 error envelope and a WWW-Authenticate Bearer challenge with realm mcp.
The docs home loads widget.kapa.ai and renders a visible Ask AI button with id kapa-button inside the kapa-widget-container shadow root.
Show the score
Paste this into a readme:
[](https://docsforagents.com/reports/doppler-docs-ai-agent-readiness/)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.
Related reports
- Recall.ai Grade B- · 80.0%
Put another docs site through the battery.
Nominate a docs site