aeokit runtime
The headless API and worker for prompts, answers, citations, mentions, competitors, visibility metrics, and opportunities.
Explore the runtimeRuntime · Agent · Skills
Know why AI answers mention—or miss—your brand.
aeokit helps you observe what answer engines say, investigate the evidence behind it, and make verifiable improvements—from first prompt to final patch.
Open source · Local-first · Every metric has receipts
Give your agent an AEO workflow
npx aeo-skills@latest add aeo-audit --to agents
# Then ask your agent:
$aeo-audit Audit this site for a buyer questionThe aeokit ecosystem
Use one layer. Or connect all three.
Start with a portable workflow, run a focused local agent, or build directly on the headless runtime. Every layer favors inspectable evidence over opaque recommendations.
aeokit agent
A local-first agent that plans, investigates, and improves answer-engine visibility while keeping evidence and changes in view.
Meet the agentaeokit skills
Portable audit, improve, and observe workflows for Codex, Claude Code, ChatGPT, Cursor, Copilot, Gemini CLI, and aeokit agent.
Browse the skillsSUPPORTED PROVIDERSOpenAIAnthropicOpenRouterGoogle AI Overviews
Evidence in practice
Every metric has receipts.
The interactive example below shows the evidence model in action. Visibility percentages are useful only when you can inspect the prompt, provider, answer, mention decision, and citation behind them. aeokit keeps the evidence attached and makes it available to apps, agents, CLI tools, and integrations.
Visibility
How often answer engines mention your brand versus cite one of your domains.
Mention rate75%
Citation rate50%
Brand mentions63
Successful runs84
| Prompt | Mention rate | Citation rate | Evidence | Providers | Last run |
|---|---|---|---|---|---|
| What are the best open-source AI visibility tracking tools? | 75% | 50% | 21 mentioned · 14 cited / 28 | OpenAIAnthropic | 2 minutes ago |
| Which tools help brands improve visibility in AI answers? | 75% | 50% | 21 mentioned · 14 cited / 28 | OpenAIAnthropic | 8 minutes ago |
| Compare self-hosted AEO platforms for a small marketing team. | 75% | 50% | 21 mentioned · 14 cited / 28 | OpenAIAnthropic | 14 minutes ago |
VisibilityHow often answer engines mention your brand versus cite one of your domains.
Explore How it works
From prompts to proof in three steps.
Define the market
Add your brand, aliases, owned domains, and the competitors that matter.
Run real questions
Choose providers, schedule prompts, and preserve each grounded answer exactly as returned.
Measure and inspect
Track visibility, share of voice, citations, and costs—then open the underlying run.
Shipped today
Visibility tracking without the black box.
The runtime stores the observations. The agent investigates and acts. The skills make repeatable AEO practice portable across the tools your team already uses.
Visibility trends
Measure the share of successful answers that mention your brand over 7, 30, or 90 days.
Share of voice
Compare your mentions with every competitor you choose to track.
Citation inventory
See the URLs and domains answer engines use as evidence, including your owned pages.
Raw answer history
Open every provider response, mention decision, citation, error, and timestamp behind a metric.
Prompt scheduling
Run the questions that matter on demand or on a recurring worker schedule.
Provider cost visibility
Show exact upstream request costs when a provider reports them—without invented estimates.
Bring your own keys
Provider credentials stay server-side and are never sent to the browser or stored in PostgreSQL.
Self-hosted by default
Run the headless API, worker, and PostgreSQL locally with Docker Compose.
Coverage
Provider coverage, stated precisely.
aeokit uses documented APIs and web-search capabilities. It does not claim that an API response is identical to a consumer chat interface.
See provider detailsOpenAIResponses APINative web search
AnthropicMessages APINative web search
OpenRouterOnline model routingModel-dependent web access
Google AI OverviewsDataForSEO SERP APIGoogle Organic Live Advanced
Why open source
Your visibility data should not be trapped in another dashboard.
Inspect the scoring logic, store raw answers in your database, choose your providers, and customize the workflow. The runtime is AGPL-3.0 licensed and designed to run on infrastructure you control.