Docs

REST API

Every endpoint for scan, batch, billing, and more.

All docs Start free

What is a content quality gate?

dokeo.co/blog/what-is-a-content-quality-gate

WARNING

76overall

77SEO

86AEO

66GEO

Fix before it ships12 of 19 checks pass

internal_linkingNo contextual links found (internal or external)

Fix

slopquestion-then-answer pattern found 2 time(s)

Fix

We gate our own blog. Scored 17 October 2026 with the same engine you get, glossary rules, balanced strictness.

REST API

Base URL: https://api.dokeo.co. Every endpoint is a JSON POST or GET. Bearer-token auth, HMAC-verified, rate-limited.

Authentication

Pass your tenant API key as a Bearer token. Create one while signed in at /account/api-keys, which calls POST /api/v1/account/keys.

bash

# All requests need an Authorization header with a Bearer token
curl https://api.dokeo.co/api/v1/scan \
  -H "Authorization: Bearer dk_your_tenant_key" \
  -H "Content-Type: application/json" \
  -d '{
    "content": "Your blog post here...",
    "content_type": "blog"
  }'

Response shape

Every scan returns the same envelope. verdict is the worst of five severities, most to least severe: BLOCKER (ships never), FAIL, WARNING, OPTIMIZATION (a suggestion, never blocks), or PASS. scores.overall is 0-100. checks_detail[] has one row per check with its own severity, reasons, and metrics.

json

{
  "slug": "my-post",
  "verdict": "WARNING",
  "content_type": "blog",
  "word_count": 1245,
  "report": {
    "scores": {
      "overall": 87,
      "seo": 92,
      "aeo": 81,
      "geo": 88
    },
    "meta": {
      "pass_count": 6, "blocker_count": 0, "fail_count": 0,
      "warning_count": 1, "optimization_count": 0
    },
    "checks_detail": [
      {
        "check": "structure",
        "status": "PASS",
        "severity": "PASS",
        "reasons": ["1245 words, in target band"],
        "metrics": { "word_count": 1245, "h2_count": 6 }
      }
    ],
    "action_items": [
      { "priority": "recommended", "action": "Add an FAQ section for AEO" }
    ],
    "inline_issues": [],
    "evaluation": {
      "content_hash": "sha256:9a1f...",
      "decision": { "verdict": "WARNING", "overall": 87, "findings": [["aeo", "WARNING"]] },
      "decision_hash": "sha256:3f9c..."
    }
  }
}

Endpoints

POST/api/v1/scan

Score inline content. 1 credit per call.

Field	Type	Notes
content	string	Required The post body
content_type	string	Default: blog One of 17 types
title	string	Optional Post title
slug	string	Optional Post slug
primary_keyword	string	Optional Main SEO keyword
strictness	string	Default: balanced strict \| balanced \| permissive

POST/api/v1/scan/replay

Re-run an earlier scan of your workspace and check the decision is identical. Free. Returns match, the inputs that changed, and both outcomes. Dokeo does not keep your text, so send it again.

Field	Type	Notes
content	string	Required The same text that was scanned
decision_hash	string	Required report.evaluation.decision_hash from the earlier scan

POST/api/v1/scan/url

Fetch + score a URL. SSRF-guarded (blocks private/loopback IPs).

Field	Type	Notes
url	string	Required https URL
content_type	string	Default: blog
strictness	string	Default: balanced

POST/api/v1/scan/blog

Score every post on a blog (discovered via RSS/sitemap).

Field	Type	Notes
url	string	Required Blog index / RSS / sitemap URL
limit	int	Default: 10 Max posts to scan
content_type	string	Default: blog

POST/api/v1/campaign/qa

Pre-send QA for an email sequence (1 to 25 emails). Runs the email checks on each message and says what it would block. Shadow mode (default) never holds a send for quality, so you can compare it with what you actually send for a few weeks. Dokeo never sends mail. Uses 1 credit per email. Returns send, action (send | hold | skip), would_block_steps and, per email, verdict, score, findings and a decision_hash you can replay. Recipients on your do-not-contact ledger are dropped from send_to in every mode. sequence_findings lists schedule problems (quiet hours, steps too close together or out of order) and steps whose voice drifts from the rest of the sequence; they hold a send in enforce mode only when they reach fail_on.

Field	Type	Notes
campaign	string	Required A name for the sequence
emails	object[]	Required Each has step, subject and body, and optionally send_at (ISO 8601)
mode	string	Default: shadow shadow \| enforce (enforce returns send: false when anything would block)
fail_on	string	Default: fail none \| blocker \| fail \| warning \| optimization
min_score	number	Default: 0 A score under this blocks, whatever the verdict
strictness	string	Default: strict strict \| balanced \| permissive
footer	string	Optional What your sender appends, such as the unsubscribe line. Scored with each email
recipients	string[]	Optional Up to 1000 addresses. Returns send_to, suppressed and invalid_recipients
timing	object	Optional recipient_timezone (IANA name), quiet_start_hour (20), quiet_end_hour (7), min_gap_hours (24). Applied to steps that have send_at

POST/api/v1/campaign/suppressions

Record a refusal on your do-not-contact ledger. Append-only: the wording is kept as you give it, a second entry for the same address keeps the first, and no model is involved. Plus-tags are ignored, so sam+news@acme.com is covered by sam@acme.com.

Field	Type	Notes
target	string	Required An address, or @domain.com for a whole domain
reason	string	Optional Up to 200 characters
source_text	string	Optional What they wrote, kept as evidence

GET/api/v1/campaign/suppressions

Everyone currently suppressed for your workspace, newest first, with the original wording.

POST/api/v1/campaign/suppressions/anonymise

For an erasure request: replace a suppressed address with its SHA-256 digest and clear the stored wording. The person stays suppressed without Dokeo holding their address. Needs a signed-in admin session.

Field	Type	Notes
target	string	Required The address or domain to anonymise

POST/api/v1/campaign/suppressions/lift

Remove a suppression. Needs a signed-in admin session (API keys cannot) and a reason. The history stays.

Field	Type	Notes
target	string	Required The address or domain to lift
reason	string	Required Why it is safe to mail them again

POST/api/v1/builder/research

Gate 1: research keyword + N tools. LLM does web search + per-tool lookup.

Field	Type	Notes
primary_keyword	string	Required
audience	string	Required e.g. 'B2B event teams'
tool_count	int	Default: 6 2-15
house_product	string	Required Your brand to lead with
llm_provider	string	Default: auto auto \| claude \| gpt \| mock
mock	bool	Default: false Use fixtures, no LLM

POST/api/v1/builder/generate

Gate 2: write the draft. Returns draft_md + QA report + editorial score.

Field	Type	Notes
research_bundle	object	Required Output of /builder/research
format	string	Default: listicle listicle \| comparison \| alternatives \| buyers_guide
llm_provider	string	Default: auto
mock	bool	Default: false

GET/api/v1/billing/entitlement

Current tenant + tier + credit balance + usage this period.

POST/api/v1/checkout/create-session

Create a checkout session for a paid tier. Requires a signed-in session (Bearer JWT).

Field	Type	Notes
tier	string	Required growth \| pro \| scale
cycle	string	Default: monthly monthly \| annual

GET/api/v1/catalog/engines

List the supported engines (SEO/AEO/GEO) and their check names.

POST/api/v1/admin_ext/auto_fix

One-click AI rewrite for a failed check (configured writing model). Requires a Growth+ plan; consumes from the rewrite allowance, never from scans.

Field	Type	Notes
content	string	Required The post body
check_name	string	Required The failing check
reasons	string[]	Optional From check.reasons[]
metrics	object	Optional From check.metrics{}

GET/openapi.json

Full OpenAPI spec at https://api.dokeo.co/openapi.json. Use `dokeo openapi --out spec.json` to dump.

Errors

Code	Meaning
401	Missing or invalid Bearer token
402	Out of credits. Upgrade or wait for refill.
404	Path or scan result not found.
422	Validation error (missing field, wrong content type).
429	Rate limit exceeded. Check X-RateLimit-Reset header.
502	Upstream (payment or search providers) unavailable.
503	Server missing required configuration.

CLI →

Same API, in your terminal

MCP server →

Call the gate from Claude / Cursor

OpenAPI spec →

Raw JSON for codegen

REST API

Authentication

Response shape

Endpoints

Errors

Related