Viralmaxing skills for agents
The connector gives an agent tools. The skills explain how to use them: where to start when breaking down an account, how research differs from reading the plan, which calls cost energy, and when to ask before spending it.
The repository is public and generated from the product's own source, so it cannot drift from it: Viralmaxing/skill.
Install
npx skills add Viralmaxing/skill
One command installs both the skills and the MCP server: the repository follows the Agent Plugins spec: skills/ for the skills, mcp.json for the connector. You do not need to add the server separately afterwards.
If you only want the connector, without the skills:
claude mcp add --transport http --scope project viralmaxing https://api.viralmaxing.com/api/mcp
What is inside
| Skill | What it covers |
|---|---|
viralmaxing-analytics | Metrics for tracked accounts: views, engagement, movement over time |
viralmaxing-research | Comparison against competitors, discovery of new accounts and ideas |
viralmaxing-video | Transcript and metric detail for one video |
viralmaxing-plan | The content plan: take a video into work, write and move a scenario |
viralmaxing-automations | Instagram comment-to-Direct funnels |
The split is deliberate: an agent loads only the skill a request needs instead of spending context on the rest.
Sign-in and money
The skills carry no secrets. Authorization goes through OAuth in the browser, same as the connector itself. See MCP quick start.
Paid actions never run silently: the tool quotes its price and waits for confirmation. What costs what is in the energy reference.