Skip to main content

Stunt Double Index

The independent ranking of how real AI agents experience the world's websites across discovery, information retrieval, task completion, delegated access and more.

Agent score▲ 1.8
77.3/100Good
Embed
Rank
Top 51%
#267
SaaS rank
Sector average 71.3
#94 of 216
Benchmark runs
3 providers
7
Where this site sits
SaaS average 71.3
+6.0
Market average 72.5
+4.8

Category breakdown

Strongest is Accuracy; weakest is Market ranking.

Brand awarenessCan agents recognize you exist?
96
▲ 8.2
DiscoveryWill they pick you?
67
▼ 1.4
Information retrievalCan they read your site?
75
▼ 1.2
Market rankingWhere do you sit in the lineup?
53
AccuracyDo they tell the truth about you?
97
▲ 8.4
Task completionCan an agent complete a task on behalf of a user?
63
▼ 2.0
Delegated accessDo you let agents in, safely?
89
▼ 0.2
Contact & communicationCan an agent reach a human?
67
▼ 3.9

Score over time

Every run against the SaaS and market medians.

▲ 1.8

2 runs since 29 Sept 2026, ▲ 1.8 overall.

  • This site
  • SaaS median
  • Market median
708090100MayJunJulAugSept
Each category over time
  • Brand awareness96▲ 8.2
  • Discovery67▼ 1.4
  • Information retrieval75▼ 1.2
  • Market ranking530.0
  • Accuracy97▲ 8.4
  • Task completion63▼ 2.0
  • Delegated access89▼ 0.2
  • Contact & communication67▼ 3.9
Every run, as a table
RunThis siteSaaS medianMarket median
30 Sept 202677.375.876.3
29 Sept 202675.575.876.3

Medians are by month: each domain counts once, at its newest run that month. Dates are UTC. Latest run 30 Sept 2026.

Agents

ChatGPT Agent does best. Each provider's model and every session behind its score.

70.5
How these are scored

The score beside each provider is its session quality over a rolling 30 days, from at least 10 sessions per provider.

Scores run 0 to 100. Each is one agent on one task, graded against that task’s fixed rubric, so no agent marks its own work. The median under a score is the sector median: the middle score for the same category, or the same provider, across the other SaaS sites the Index has sent agents to.

  • Excellent85+
  • Good72-84
  • OK58-71
  • Poor40-57
  • Critical<40

Latest run 30 Sept 2026 (UTC). Anyone signed in can re-run it from 1 Oct 2026, 00:20 UTC. The owner can re-run it any time from their dashboard.

Claudeclaude-sonnet-5
07 tasks69.5
Latest run average69.5median 90.4
  • Market ranking
    market-ranking-ecom-v2 · 11s
    43median 66 Done
  • Delegated access
    delegated-access-ecom-v1 · 1m 34s
    96median 91 Done
  • Task completion
    journey-ecom-v1 · 1m 57s
    0median 86 Done
  • Discovery
    journey-ecom-v1 · 1m 57s
    0median 91 Done
  • Information retrieval
    research-ecom-v1 · 2m 37s
    36median 76 Done
  • Brand awareness
    research-ecom-v1 · 2m 42s
    98median 95 Done
  • Accuracy
    research-ecom-v1 · 2m 37s
    100median 64 Done
ChatGPT Agentgpt-5-mini
08 tasks70.5
Latest run average70.5median 84.6
  • Market ranking
    market-ranking-ecom-v2 · 18s
    43median 66 Done
  • Contact & communication
    support-ecom-v1 · 1m 59s
    26median 79 Done
  • Delegated access
    delegated-access-ecom-v1 · 2m 39s
    94median 91 Done
  • Task completion
    journey-ecom-v1 · 4m 32s
    49median 86 Done
  • Discovery
    journey-ecom-v1 · 4m 30s
    26median 91 Done
  • Information retrieval
    research-ecom-v1 · 10m 55s
    62median 76 Done
  • Accuracy
    research-ecom-v1 · 10m 56s
    100median 64 Done
  • Brand awareness
    research-ecom-v1 · 10m 56s
    100median 95 Done
Geminigemini-3.8-flash
08 tasks69.4
Latest run average69.4median 88.1
  • Market ranking
    market-ranking-ecom-v2 · 12s
    36median 66 Done
  • Contact & communication
    support-ecom-v1 · 1m 21s
    36median 79 Done
  • Delegated access
    delegated-access-ecom-v1 · 3m 23s
    93median 91 Done
  • Task completion
    journey-ecom-v1 · 5m 24s
    49median 86 Done
  • Discovery
    journey-ecom-v1 · 5m 24s
    58median 91 Done
  • Accuracy
    research-ecom-v1 · 7m 11s
    85median 64 Done
  • Brand awareness
    research-ecom-v1 · 7m 12s
    88median 95 Done
  • Information retrieval
    research-ecom-v1 · 7m 13s
    76median 76 Done

Where agents get stuck

5 issues, worst first.

The public summary. Verified owners can replay every session in full.

MediumMarket ranking
No price visible to non-JS clients. Agents may report "pricing not disclosed".
MediumTask completion
No visible primary call-to-action in the server-rendered HTML. Agents have nothing concrete to click on behalf of a user.
MediumTask completion
No mention of delegated auth primitives (passkey, OAuth, MCP). Agents must drive a full browser session.
MediumDelegated access
No OAuth/OIDC discovery metadata. Agents cannot bootstrap a scoped, revocable session; any delegation falls back to password sharing.
Show 1 more issue
LowMarket ranking
No review or aggregateRating schema. Agents have no signal for ranking vs peers.

Want to find these in your own flows before an agent does? Read how to test your product for AI agent users.

Technical readiness

The protocol baseline: robots, sitemaps, llms.txt, structured data and more.

63/100

Protocol baseline by Is Agentic

Score
63/ 100
Important blockers remain
Essential
44.8/ 80
5 of 11 passing
Recommended
14.4/ 20
10 of 18 passing
Bonus
+4.2
19 positive signals
Fail
OpenAPI spec publishedessential
No OpenAPI/Swagger specification found
Fix: Publish an OpenAPI (Swagger) specification at /openapi.json or /api/openapi.yaml. This is how agents understand your API surface automatically.
Fail
Scoped permissionsessential
No declared OAuth scopes, security schemes, or scoped-permission documentation found
Fix: Declare scoped API permissions where machines can read them: named OAuth scopes in your OpenAPI security schemes, or scopes_supported in RFC 9728 protected-resource metadata. Prose descriptions of roles help humans, but agents need the machine-readable declaration to request least-privilege access.
Fail
JSON error responsesessential
API does not return JSON error responses (or no API detected)
Fix: Return structured JSON error responses with error codes, messages, and resolution hints. Agents can't parse HTML error pages.
Show 11 more issues
Fail
Markdown content negotiation (acceptmarkdown.com)essential
Homepage https://index.stuntdouble.io does not meet the Markdown negotiation requirements: Accept: text/markdown returned text/html; charset=utf-8; Vary header missing Accept (got "rsc, next-router-state-tree, next-router-prefetch, next-router-segment-prefetch")
Fix: Enable Markdown negotiation on the scanned homepage. Supporting it only on /docs or a separate .md URL does not satisfy this check. Requests with Accept: text/markdown must receive a nonempty Markdown body with Content-Type: text/markdown and Vary: Accept. Keep serving HTML for Accept: text/html. Adding Vary alone does not create a Markdown response. Verify both with `curl -sS -L -i -H 'Accept: text/markdown' https://yourdomain.com/` and `curl -sS -L -i -H 'Accept: text/html' https://yourdomain.com/`. Check the final response headers and body: Markdown with Vary: Accept for the first request, HTML for the second.
Fail
Agent instruction / when-to-userecommended
No agent instruction file with when-to-use guidance found
Fix: Tell agents when to reach for you: add a 'when to use this' section to your llms.txt (or a dedicated agent-instructions file) that names your best-fit use cases and how an agent should call you. Be specific about the jobs you are right for - generic marketing copy does not read as guidance.
Fail
API schema complexity analysisrecommended
No API schema detected
Fix: Make your API spec self-describing: a unique operationId and a description on every operation, typed parameters, and response schemas. For GraphQL, a fully typed schema with a documented cost or rate limit reads best.
Fail
Function calling compatibilityrecommended
No API spec found - function calling requires discoverable endpoints
Fix: Ensure API endpoints have unique operation IDs, typed schemas, and descriptions compatible with LLM function-calling formats.
Partial
Agent-friendly 404sessential
The nonexistent path https://index.stuntdouble.io/__ora-404-probe-b7jllor5 correctly returns HTTP 404. Partial credit: no Markdown error body was detected.
Fix: Keep the correct HTTP 404 status. The remaining requirement is a Markdown error body when agents request Accept: text/markdown. Include at least 20 characters explaining the error and a link to your docs, sitemap, or llms.txt. Verify with `curl -sS -L -i -H 'Accept: text/markdown' https://yourdomain.com/some-path-that-does-not-exist`. Check both the final 404 status and the Markdown body with Content-Type: text/markdown. Checking the status alone does not verify the missing requirement.
Partial
Content without JavaScriptessential
3362 chars with H1, but 2.5% content ratio is below the 5% target
Fix: Serve at least 500 characters of meaningful homepage content in raw HTML. Add a clear H1, keep deeper heading levels sequential, and remove excessive non-content markup.
Partial
Developer resource discoverabilityrecommended
Agent struggled to find developer resources by name - only 1 result(s) out of 10. Not found in this search sample: OpenAPI spec, SDK documentation
Fix: Check whether your developer resources (API docs, OpenAPI spec, auth docs, developer portal, MCP server, SDK documentation) surface in name-based searches. If they do not, use predictable URLs, link them in llms.txt, and include your product name in page titles and headings. This result reflects one search sample.
Partial
Agent onboarding frictionrecommended
Onboarding signals described but not verified live: free tier available, self-serve key generation, sandbox/test environment
Fix: Offer a free tier or trial, self-serve API key generation, and a sandbox environment. Agents can't fill out 'contact sales' forms.
Partial
Organization schema completenessrecommended
Organization schema found but missing: address
Fix: Add Organization JSON-LD that includes both contactPoint (with email/phone and contactType) and address (PostalAddress). This lets AI verify your business legitimacy and answer contact queries.
Partial
JSON-LD structured datarecommended
JSON-LD Organization found - missing description on this entity
Fix: Add description to the Organization JSON-LD entity for full score.
Partial
MCP server / manifestrecommended
MCP server detected at https://app.stuntdouble.io/api/mcp; it requires authentication. Presence is verified (5/6), but Ora could not inspect its tools or verify access scopes without credentials.
Fix: Keep authentication enabled. Verify tool listing with an authorized MCP client. This unauthenticated scan cannot establish whether protected tools are missing or unusable.
29 checks · scanned 29 Sept 2026

Related domains

index.stuntdouble.io is a subdomain of stuntdouble.io. Agent experience can differ sharply across subdomains.

Embed

Put this score on your site

A live badge for your footer, docs or README. It updates after every run and links back to this report.

index.stuntdouble.io Stunt Double Index agent score badge