Map every HTTP API this box serves and calls, and catch it when one changes

Every HTTP API the host serves and calls, read from live traffic with TLS included: the process on each end, each endpoint's status codes and p50/p95, and the JSON shape of its requests and responses, learned from real bodies. Then it keeps watching, and a field that stops arriving, a type that changes, a new 500 or a p95 that doubles lands in a drift feed with the endpoint and the detail.

Modeled on yeet-src/httpscope

Set up httpscope (https://github.com/yeet-src/httpscope) on this Linux machine. Its README has a "Have an agent set it up" prompt: follow it.

Work out where I'll open the page from (this machine, or another computer through SSH or a VM's port forwarding) and make sure it's reachable there. Give me the exact address to type. If `localhost` might resolve to IPv6 and reach something else, give the 127.0.0.1 form.

`npm install` and `make bpf` run code from the repo; if your permissions block them, tell me the exact command to run myself.

Done means: the page loads for me and shows real HTTP traffic. If the machine is quiet, send a couple of harmless requests of your own (not to my services) to prove it. Explain what I'm looking at in plain language.

The API docs nobody wrote.

The spec says what the service was meant to return. What it returns today, to whom, and what it calls in turn lives in nobody's head and no file. When a field quietly stops arriving after a deploy, the first sign is a client breaking somewhere else.

  • An OpenAPI spec that drifted from the running service months ago
  • An endpoint nobody remembers that still gets traffic
  • A missing field found by the client that broke, not by the service that dropped it

One prompt, this much work. Every step on your own hosts.

  1. 1Tap every interface for plaintext and the TLS libraries for HTTPS, with no change to any app
  2. 2Reassemble each connection into HTTP/1.x and HTTP/2 transactions
  3. 3Group them into services and endpoints, with the process on each end
  4. 4Learn each endpoint's request and response shape from its real bodies
  5. 5Flag drift: a field missing, a type changed, a new status code, a p95 that doubles
  6. 6Serve it as six live pages for people and a GraphQL API for agents

A live map of your APIs, and a note when one moves

Within a minute, a table of every API the host serves and calls with who is on each end, each endpoint's statuses, latency and shape, and a drift feed that names the endpoint and the change the moment it happens. Nothing in your services changes.

More prompts in Debug your APIs

  • API alerts

    DockerKubernetesSlack

    List the APIs this machine serves and calls from live traffic, then alert me in Slack when one returns 5xx, its 4xx jump, or its port goes quiet.

    See what it builds

  • HTTP top

    DockerKubernetes

    Build me `top` for the HTTP endpoints on this box: every request, ranked live, no proxy.

    See what it builds
← Every prompt

Install yeet, then paste a prompt

curl -fsSL https://yeet.cx | sh

Linux only  |  Manual install  |  Read the docs

Sign up