Skip to main content
Give your agent — Claude Code, Cursor, Codex, or anything else with a shell — the ability to search, inspect and run paid API endpoints under one Key.
1

Install the Skill

Paste this line into your agent’s chat:
The agent reads the Skill, installs the glasser CLI, and learns when to use it, how to read a price before spending, and how to retry without paying twice.
2

Approve the login

The agent runs glasser login and shows you a URL and an 8-character code. Open the URL, check that the code on the page matches the one in the terminal, and click Approve.A Key scoped to your Workspace lands on the machine. You never copy or paste it, and the agent reads the Workspace name back to you as confirmation. New here? The login page creates your account and Workspace on the way.
3

Try a prompt

The agent now searches the data sources, inspects the price, runs the endpoint, and reports the result and the charge. Try one of these:
Using Glasser, enrich a person. Ask me for their name and current company if you don’t know them, then pull their work history, education and public profiles, and tell me what the call cost.
Do a competitor research report using Glasser. Ask me which company to research, then pull its firmographics, recent news and web presence, and compare it with two competitors.
Using Glasser, find companies hiring for a role. Ask me for the role and location, then list open postings with company, title and posting date.
Using Glasser, research keywords for a niche. Ask me for the topic and the market, then return search volume and related keyword ideas, ranked.

​
What the Skill promises

The Skill is the agent’s working agreement, not just an installer. It tells the agent to inspect before it runs, to name the per-call price before any bulk or looped work, to start with small volume parameters, to retry an ambiguous failure with the same idempotency key so you are never charged twice, and to report both the run status and what the provider said. Read it yourself at glasser.ai/SKILL.md.

​
No shell?

If your agent cannot run commands — a web chat, a hosted runtime — connect the MCP server instead. Same data sources, same Key, same rules.