◎ Discry Score
apollo graphql.com
devtools · API
B
0 / 100
DISCOVERY0
COMPREHENSION0
Category leader: 96 (A)
Discry your API →
DEVTOOLS · RANK #16 OF 39

Apollo GraphQL is Good to agents.

Discry independently scored how well an AI agent can discover and understand the Apollo GraphQL API from what’s public — not whether it’s usable. Below: every signal we checked, what’s costing the score, and what to change.

Discry your API — freeView the docs ↗

SCORED UNDER RUBRIC 1.2 · A full re-launch under Discry Score 2.5 — a new behavioral instrument, not comparable to these scores — is in progress.

Discovery

45% of score · 62/100
OpenAPI specA machine-readable OpenAPI/Swagger spec agents can parse.Fail
llms.txtAn llms.txt index that points agents to the docs that matter.Pass
llms.txt qualityThe llms.txt is focused, current, and well under the size budget.Partial
llms-full.txtA full-text bundle agents can load in one request.Fail
AGENTS.mdAn AGENTS.md that tells coding agents how to build on the API.Pass
.well-known/mcp.jsonA discoverable MCP manifest at a well-known path.Fail
MCP registryThe API is listed in a public MCP registry.Pass
robots.txt AI directivesrobots.txt allows (or explicitly guides) AI crawlers.Pass
SitemapA sitemap so agents can enumerate the docs surface.Pass

Comprehension

55% of score · 91/100
Task-oriented descriptionsEndpoints described by what they accomplish, not just their shape.Pass
Realistic examplesRunnable, real-world request/response examples.Pass
Multi-step workflowsDocs that chain calls into complete jobs an agent can follow.Pass
Error-recovery guidanceDocumented failure modes and how to recover from them.Partial
Answer-first formatThe answer leads; preamble does not bury it.Pass
Capability boundariesClear limits — what the API can and cannot do.Pass
Naming consistencyConsistent, predictable naming across endpoints.Pass
Heading hierarchyClean heading structure agents can navigate.Pass
Markdown docsDocs available as clean markdown, not JS-rendered HTML only.Pass
Token efficiencyDocs are concise enough to fit an agent context window.Partial

What we found

  • An agent would discover Apollo easily and be explicitly welcomed: robots.txt carries a Content-Signal directive permitting ai-train and ai-input, and every docs page is served as clean markdown (.md endpoints) — so an agent gets structured content without HTML parsing.
  • Apollo is unusually agent-native. There is an official Apollo MCP Server (listed on PulseMCP), a dedicated apollographql/skills Agent Skills repo, hosted GraphOS MCP Tools at mcp.apollographql.com, and AGENTS.md/CLAUDE.md in primary repos — far more agent tooling than a typical API ships.
  • Comprehension is near-exemplary: task-oriented descriptions, realistic copy-pasteable examples (typed GraphQL mutations, real CLI commands), multiple end-to-end multi-step guides, and inline error-recovery hints (e.g. 'If you receive Error: Not connected to GraphOS, ensure you've included your APOLLO_KEY').
  • The biggest discovery gap is structural, not a quality failure: there is no OpenAPI spec because Apollo is GraphQL. The native equivalent (an introspectable schema — the Platform API is itself a public graph) exists, but Discry's highest-weighted discovery check can't credit it, which caps the discovery score.
  • The llms.txt is comprehensive but heavy (~139KB / ~35K tokens). It works as an excellent sitemap for agents, but is too large to load wholesale into a context window, and no llms-full.txt exists as a single-pull corpus.

What to change

Prioritized by impact on discoverability. You (or your docs platform) deploy these — Discry never touches your API.

  1. 01Add an llms-full.txt (currently 404) so agents can pull one comprehensive corpus without crawling, and slim the llms.txt index toward an API-focused subset (<50KB) to improve token efficiency.
  2. 02Publish a machine-discoverable schema entry point for the Platform API — a documented public introspection/SDL endpoint or a .well-known pointer — so agents can locate the GraphQL 'spec' where they'd otherwise expect an OpenAPI URL, recovering the OpenAPI-shaped discovery gap.
  3. 03Add /.well-known/mcp.json advertising the hosted GraphOS MCP Tools endpoint (mcp.apollographql.com) so agents auto-discover the MCP server instead of finding it only via blog posts or third-party registries.
  4. 04Consolidate error-recovery guidance into a dedicated per-error 'what to do next' reference covering the top publish, auth, and schema-check failures, rather than leaving recovery hints scattered inline in the quickstart.

Execution coverage · INFORMATIONAL, UNSCORED

Whether an agent can actually complete a call and recover from errors is the deeper Audit layer — documented here, but not part of the Discry Score.

API key✓ Error format documented✓ Rate limits documented✓ Pagination documented✕ Idempotency documented

Execution characteristics are well-documented for a GraphQL platform: API-key auth (graph/user/subgraph keys via APOLLO_KEY), GraphQL-standard JSON error responses, explicit publish rate limits (400 req/min per graph), and cursor/offset pagination patterns in the Apollo Client docs. Idempotency keys are not a documented concept.

See your own Discry Score.

Drop your API docs URL. See what an agent sees — in 60 seconds, free.

Discry your API — free