Reference

Docs

Four tools — assess_ad_creative, audit_campaign_structure, analyze_campaign_performance, strategize_targeting. Two front doors: MCP (Streamable HTTP) and REST. One key for both — create it in the dashboard.

Connect over MCP

Claude Code

claude mcp add --transport http spendict https://spendict.com/api/mcp \
  --header "Authorization: Bearer spd_live_YOUR_KEY"

Claude Desktop · Cursor · Windsurf

Add to your MCP config (claude_desktop_config.json / .cursor/mcp.json):

{
  "mcpServers": {
    "spendict": {
      "url": "https://spendict.com/api/mcp",
      "headers": { "Authorization": "Bearer spd_live_YOUR_KEY" }
    }
  }
}

n8n

Use the MCP Client Tool node — Server Transport = HTTP Streamable, Endpoint = https://spendict.com/api/mcp, header credential Authorization: Bearer spd_live_YOUR_KEY.

The four tools

Every tool is exposed over MCP at /api/mcp and mirrored one-for-one over REST. Same auth, same verdicts.

assess_ad_creative	POST /api/v1/assess	run / fix_first / kill on an ad creative
audit_campaign_structure	POST /api/v1/audit-campaign	sound / fix_first / restructure on a build
analyze_campaign_performance	POST /api/v1/analyze-performance	healthy / fix_first / kill_or_rebuild / insufficient_data on live metrics
strategize_targeting	POST /api/v1/strategize-targeting	a generated targeting strategy (structure-validated)

assess_ad_creative — gate the creative

The flagship. Judges an ad before a cent is spent and returns a deterministic verdict.

curl -X POST https://spendict.com/api/v1/assess \
  -H "Authorization: Bearer spd_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "ad_copy": {
      "primary_text": "Still paying an agency $3k/mo for ads a robot writes better?",
      "headline": "Fire your ad agency"
    },
    "platform": "meta",
    "format": "static",
    "product_context": "AI Meta-ads automation tool, $99/mo",
    "target_audience": "DTC founders spending $2-20k/mo on Meta ads"
  }'

Input

ad_copy.primary_text	string · required	The main ad body or video script.
ad_copy.headline	string	Optional headline.
ad_copy.description	string	Optional description / subtext.
creative_url	https URL	Optional image or video-frame URL — enables visual assessment.
platform	meta \| tiktok \| google \| linkedin \| youtube · required	Assessment is platform-specific.
product_context	string · required	What's being sold, the offer, the key value.
target_audience	string	Who it's for, if known.
format	video \| static \| carousel	Defaults to static.

The verdict

Scores are 0–100 across seven dimensions (hook, angle, clarity, audience_resonance, platform_fit, cta, compliance_safety) plus an authoritative, deterministic launch_recommendation:

run — worth spending on: overall ≥ 75, hook ≥ 60, no hard gate tripped.
fix_first — a real spine but a specific fixable problem; the fixes are the path to run. A dead hook or a likely policy rejection always caps an ad here.
kill — structurally broken (two or more failing fundamentals); faster to regenerate than to fix.

{
  "overall_score": 78,
  "launch_recommendation": "run",
  "dimension_scores": {
    "hook": 78, "angle": 80, "clarity": 82,
    "audience_resonance": 80, "platform_fit": 85,
    "cta": 0, "compliance_safety": 95
  },
  "predicted_failure_mode": "No material weakness — would run as-is.",
  "prioritized_fixes": [],
  "enhancements": [
    { "issue": "No conversion path (CTA scored 0)",
      "suggestion": "Add one clear next step — optional; ships as-is." }
  ]
}

audit_campaign_structure — audit the build

Judges a campaign's structure by the correct platform's rulebook before budget flows through it. Verdict: sound / fix_first / restructure.

platform	meta \| google \| tiktok	Required.
objective	string	Required — defines 'correct structure'.
campaign_structure	string	Required — the build to audit.
total_budget / funnel_stage	string	Optional context.

analyze_campaign_performance — diagnose live

Judges real metrics against benchmarks and finds the one bottleneck. Verdict: healthy / fix_first / kill_or_rebuild / insufficient_data.

platform / objective	string	Required — set the benchmarks.
metrics	object	Required — the real campaign numbers.
linked_assessments	object	Optional — reconcile vs pre-flight prediction.
campaign_identifier	string	Optional — auto-match fallback.

strategize_targeting — design the plan

The generative tool. Given a product, budget, and objective, it returns one opinionated, executable targeting strategy — audience approach, segmentation, exclusions, a budget split with a learning-phase check, setup notes, and the single biggest risk. No score here; the server validates the structure of the returned plan.

platform / product / budget / objective	string	Required.
funnel_stage / vertical	string	Optional — axes of justified segmentation.
audience_data / geo_language	string	Optional — thin data → lower confidence.

Errors

Always JSON — { error, message }, never a stack trace.

400	invalid_input / invalid_json	Schema details; not counted.
401	missing_api_key / invalid_key	Check the Authorization header.
429	quota_exceeded	Includes upgrade_url; not a crash.
502 / 503	model_error / config_error	Not counted; retry.

Quotas

Free: 100/mo · Starter ($19): 1,500/mo · Pro ($49): 5,000/mo · Scale ($199): 25,000/mo. Every tool call counts equally. The quota check happens before inference; failed calls are refunded. Median latency is a few seconds — size your agent's tool timeout to ~30s.

Get your API key