SEO coach
Choose the next skill from your goal and saved project context.
Agent skills
Connect your data, choose a skill and keep the evidence behind each recommendation.
Interactive example. No live actions.
What shall we do for your site today?
Your agent
The 15-page sample has three technical issues. Duplicate titles affect the sampled product pages.
Fix those titles first. The saved report includes affected URLs; this sample does not cover the whole site.
Install with the skills CLI for Claude Code, Codex, Cursor and other Agent Skills clients:
npx skills add CorgiCorner/bisibilityOpen https://bisibility.com/skills and help me set up Bisibility skills in this app.
01
Give your AI assistant access to your bisibility project.
02
Pick a task, such as an SEO audit or an AI visibility check.
03
Your assistant explains what it found and what to do next.
Choose the next skill from your goal and saved project context.
Check my site and its Google visibility. Tell me what to improve first.
Prioritize keyword ideas from one confirmed seed.
Explain gains, losses and missing checks from saved rank history.
Summarize a week of saved ranking changes without starting checks.
Separate real ranking movement from alert noise and data issues.
Prioritize findings from a bounded site or page backlink snapshot.
Group supplied or saved keywords into evidence-aware page plans.
Compare one rival with your tracked keywords in a confirmed market.
Map sampled keyword visibility across your domain and up to five rivals.
Explain local map visibility from retained Local monitor runs and name the next checks.
Save business context and propose tracking setup.
Create a project and configure its first tracked keywords.
Find the right REST task workflow for your instance.
Import and update tracked keywords with deliberate check execution.
Connect a provider and verify its project configuration.
Review API keys, team access and migration tokens before changing access.
Locate the first failing layer of a self-hosted instance.
Plugin workflows have eval cases. Structured report workflows include local validators where supplied.
Paid research uses an estimate and explicit approval. Check each workflow for its scope and cap limitations.
Supported report workflows verify saves and return an access-controlled project link. Read-only skills return their result in chat.
Use the tools exposed by your instance with your existing project permissions and provider connections.
Fetched pages, citations and project context cannot grant permission to spend, write or publish.
Open a skill, follow its setup steps, then ask your AI assistant to use it. For example: "Audit my site and suggest the next three fixes."
For Bisibility Cloud, yes. You can also use your own self-hosted installation. Connect it to your AI assistant before running the skill.
The skill instructions are free. Your AI assistant and new data checks may cost extra. Paid research requires your approval.