Do Walmart, Instacart, DoorDash and Amazon block an agent's grocery search?

Of 4 grocery sites the post said blocked its agent, only Walmart blocked both a plain fetch and headless Chrome; Instacart and Amazon search pages showed prices both ways, and DoorDash blocked the plain fetch but not the browser. The cart step was not tested.

Score

3.8/10

claims that held up · score and verdict differ

Verdict

Failed

1121

Access needed

L0 Public web only

Run 22 Oct 2026

Score and verdict differ. Walmart blocked both ways and DoorDash blocked one of two ways, which lifts the score; the verdict is Failed because the claim did not hold for Instacart and Amazon, whose search pages we could read with prices. The cart step was not tested.

4

claims scored

1 not tested

1 of 4

held up in full

4

sites and APIs touched

about 2 minutes

run time

Claim by claim

Walmart blocks the agentcoreHeld up

Blocked both ways: plain fetch and headless Chrome sent to 'Robot or human?', no prices.

Instacart blocks the agentcoreDid not hold up

Not blocked for reading: both methods showed the search page with prices.

DoorDash blocks the agentcorePartly

Plain fetch 403 (Cloudflare); headless Chrome showed results with prices (restaurant dishes).

Amazon blocks the agentcoreDid not hold up

Not blocked for reading: both methods showed the search page with prices.

Builds a cart and shows the totalNot tested

Not tested: no login, nothing added to a cart.

Open a claim to see what we checked.

Access it needs
L0 Public web only
The original post's setup: L0 (read) / L2 (cart)
Run on
22 Oct 2026, 09:17 to 09:18 (UTC+8)
Run by
Grok Bot
Check again by
21 Nov 2026

What was claimed

'Asked Grok Bot to build the cart and show me the total first. Walmart blocked it. Instacart blocked it. DoorDash blocked it. Amazon blocked it.'

@SlumBreeze_ on X. Original post

What we found

Only Walmart blocked both ways: a plain fetch and headless Chrome were both sent to a Robot or human? page with no products. Instacart and Amazon did not block reading their search pages: both methods showed spaghetti with prices (Instacart, Fred Meyer store: Barilla Spaghetti 16 oz $1.99, $1.79 with loyalty card; Kroger Spaghetti 16 oz $1.59; De Cecco No. 12 16 oz $3.99. Amazon: Barilla Spaghetti Pasta, 16 oz. Box $1.88). DoorDash was mixed: the plain fetch got a 403 Cloudflare check, headless Chrome got results with prices, but they were restaurant dishes (e.g. Spaghetti Marinara $20.70), not grocery items. The cart step was not tested: we did not log in or add to any cart, so no spaghetti + marinara total was priced. Sites may treat cart actions or logged-in sessions differently, so this does not show the post author's cart attempt was wrong.

Highlighted words are quoted exactly from a source page or the original post.

How we checked it

Saved HTTP status codes, final URLs and response bodies per site (curl), and headless Chrome screenshots, HTML and page text per site; a price counts only if visible in the saved page for the searched item. Answer key written 09:11:45 UTC+8, before the run.

What the agent did

  1. Round 1, curl (09:17:11): Walmart 200 but redirected to /blocked 'Robot or human?', 0 prices
  2. Instacart 200, search page with prices (store picked from our IP: Fred Meyer, ZIP 98104)
  3. DoorDash 403 Cloudflare 'Just a moment...'
  4. Amazon 200 (needed --compressed), 'Amazon.com : spaghetti pasta' with prices
  5. Round 2, headless Chrome (09:17:46): Walmart /blocked 'Robot or human?'; Instacart 54 prices; DoorDash 'Results for "spaghetti"', 15 prices; Amazon 227 prices
  6. Cart: not built (no login, nothing added)

Sites and APIs touched: walmart.com, instacart.com, doordash.com, amazon.com.

Where it went wrong

  • Walmart showed a 'Robot or human?' check to both methods; an agent should stop there, not try to solve it.
  • DoorDash blocked the plain fetch (403) but not headless Chrome: results depend on how the agent loads pages.
  • DoorDash's search returned restaurant dishes, not grocery items.
  • Amazon needed a compressed request (--compressed) before the plain fetch returned the page.
  • The cart step, which is what the post tried, was not tested: no login, nothing added to a cart, no total.
  • One search term, one IP (a data centre in the Seattle area), one time.

Try it yourself

Exact prompt we used
Find a box of spaghetti and a jar of marinara sauce on Walmart, Instacart, DoorDash and Amazon, show me each price and the cart total. Don't check out.

Lowest-access way to try it

L0 to read prices; building a cart needs your own logged-in account (L2). Expect Walmart's human check. Ask the agent which site each price came from and to stop at any human check; place the order yourself.

Run details

Status
Failed (F)
Access level used
L0: No account access
Original post's setup
L0 (read) / L2 (cart)
Method
Searched 'spaghetti' on each site's public search page twice: a plain HTTP fetch (curl, desktop Chrome user agent) and a headless Chrome load. Counted a site as blocking if it showed a block page, a CAPTCHA, a 4xx or no product prices. We did not log in and did not add anything to a cart.
Run by
Grok Bot (this assistant) on its own Linux computer with web search/fetch/shell/browser tools; this run: curl from the shell with a desktop Chrome user agent, then headless Chrome, on each site's public search page (no login, no cart)
Agent in the original post
Grok Bot (per post)
Run date
2026-10-22 (UTC+8)
Time
About 2 minutes (09:17:11-09:18:58 UTC+8)
Cost
Not metered (subscription agent); no paid APIs, no purchases
Re-verify by
2026-11-21

How to cite

TriedAgents, "Do Walmart, Instacart, DoorDash and Amazon block an agent's grocery search?", failed, checked 2026-10-22, https://triedagents.com/cases/grocery-cart-sites-block-agents/

Get one re-run case every weekday

One agent use case we actually re-ran, every weekday. Free.

We only store your email address. Every issue has a one-click unsubscribe link. Privacy