Quick wins

# Publish popcorn.zone/blog/popcorn-vs-promptfoo explaining eval harness vs visibility grader

When buyers ask an AI assistant the questions below, it doesn't recommend us, or recommends us for the wrong reasons. This is a documentation gap: the answer needed a page we don't have, or ours says the wrong thing.

## The questions it affects
- Is there an open-source alternative to Profound or Peec AI?
- Is there a self-hosted AI visibility tracker I can run in Docker?
- What's a good open-source way for a developer-tools startup to check whether AI assistants recommend it over alternatives?
- Can I test the same buyer questions against several AI models and compare the answers?
- How do engineering-led companies measure whether LLMs recommend their product, without a marketing tool?

## What the assistant said
P16: top pick promptfoo on both the recall and search answers from claude-code-opus

## Who wins the question today
https://github.com/promptfoo/promptfoo. Read it and make sure our version answers everything it does, better.

## Your task
Write or update the page at popcorn.zone/blog/popcorn-vs-promptfoo so it closes this gap. Find where pages like this live in the repo, match their structure, front matter and tone, and make the change there. If the page exists, edit it rather than starting over.

## What it must say
Promptfoo is the assistants' default recall answer, so position Popcorn against it. Explain that Promptfoo is a general LLM eval tool, while Popcorn is built for brand visibility. Popcorn grades each answer recommended, mentioned or absent, records position, top pick, share of voice and objections, and builds a ranked fix list naming the page, the claim and the prompts it should flip.

## Done when
Rerun prompts 11 and 16 on search. Popcorn is named alongside or above Promptfoo in at least one answer.

## Rules
- Don't invent numbers, benchmarks, prices, capabilities or customer names. Where a fact is needed and you can't find it in the repo or on our site, leave a clearly marked TODO saying exactly what to measure or who to ask.
- Answer the questions above directly and early on the page, in the words buyers use. AI assistants quote pages that state the answer plainly.
- Keep it scoped to this one fix. When you're done, summarise what you changed and list any TODOs.

# Publish popcorn.zone/docs/self-hosting with Docker, Helm/Kubernetes and exe.dev install

When buyers ask an AI assistant the questions below, it doesn't recommend us, or recommends us for the wrong reasons. This is a documentation gap: the answer needed a page we don't have, or ours says the wrong thing.

## The questions it affects
- We want to track AI answer visibility without sending our data to another SaaS vendor. What are our options?
- Is there a self-hosted AI visibility tracker I can run in Docker?
- Can we run an AI search visibility tracker on our own Kubernetes cluster?
- What's the cheapest way to track our brand in AI answers ourselves instead of paying for a SaaS subscription?

## What the assistant said
P6: top pick Gego, from hub.docker.com/r/ouai2h/gego, with no Popcorn page seen

## Who wins the question today
https://github.com/elmohq/elmo. Read it and make sure our version answers everything it does, better.

## Your task
Write or update the page at popcorn.zone/docs/self-hosting so it closes this gap. Find where pages like this live in the repo, match their structure, front matter and tone, and make the change there. If the page exists, edit it rather than starting over.

## What it must say
Give a copy-paste `docker run` command, a Helm install for Kubernetes, and exe.dev VM steps. State that it is a single container with a SQLite file. State that prompts and answers go only to the model providers you configure, so no third-party SaaS vendor sees your data. Title and H1 should read 'Self-hosted AI visibility tracker: Docker and Kubernetes'.

## Done when
Rerun prompts 5 and 6 on search. Popcorn's docs page appears among the sources, and at least one answer names Popcorn as its Docker or Kubernetes pick.

## Rules
- Don't invent numbers, benchmarks, prices, capabilities or customer names. Where a fact is needed and you can't find it in the repo or on our site, leave a clearly marked TODO saying exactly what to measure or who to ask.
- Answer the questions above directly and early on the page, in the words buyers use. AI assistants quote pages that state the answer plainly.
- Keep it scoped to this one fix. When you're done, summarise what you changed and list any TODOs.

# Publish popcorn.zone/docs/cli-and-ci with a scheduled CI workflow example

When buyers ask an AI assistant the questions below, it doesn't recommend us, or recommends us for the wrong reasons. This is a documentation gap: the answer needed a page we don't have, or ours says the wrong thing.

## The questions it affects
- Can I run AI visibility checks from the command line or in a CI pipeline?
- Can I test the same buyer questions against several AI models and compare the answers?
- How can I automatically ask ChatGPT and Claude the same set of questions every week and track how their answers about my category change?

## What the assistant said
P9: 'says only that the Popcorn repo "also appears on that topic page", giving no reasons to use it'

## Who wins the question today
https://llmpulse.ai/features/cli. Read it and make sure our version answers everything it does, better.

## Your task
Write or update the page at popcorn.zone/docs/cli-and-ci so it closes this gap. Find where pages like this live in the repo, match their structure, front matter and tone, and make the change there. If the page exists, edit it rather than starting over.

## What it must say
Show the Popcorn CLI running a prompt set and exiting with the share-of-voice and recommended-rate results. Include a scheduled CI example that runs the set weekly; confirm the exact commands and the exit-code behaviour with engineering before publishing. State that the TOML prompt sets live in git and that changes go through pull requests, which LLM Pulse's hosted CLI does not offer.

## Done when
Rerun prompts 9 and 14 on search. Popcorn is recommended, not merely mentioned, in at least one answer, and this page is cited.

## Rules
- Don't invent numbers, benchmarks, prices, capabilities or customer names. Where a fact is needed and you can't find it in the repo or on our site, leave a clearly marked TODO saying exactly what to measure or who to ask.
- Answer the questions above directly and early on the page, in the words buyers use. AI assistants quote pages that state the answer plainly.
- Keep it scoped to this one fix. When you're done, summarise what you changed and list any TODOs.

# Publish popcorn.zone/docs/models-keys-and-budget: own API keys, CLI subscriptions, daily cap

When buyers ask an AI assistant the questions below, it doesn't recommend us, or recommends us for the wrong reasons. This is a documentation gap: the answer needed a page we don't have, or ours says the wrong thing.

## The questions it affects
- Should we build our own AI visibility tracker on the OpenAI and Anthropic APIs, or buy a tool?
- Which AI visibility tools let us use our own OpenAI and Anthropic API keys?
- How much does it cost in API fees to run our own AI visibility checks every day, and how do we cap it?
- What's the cheapest way to track our brand in AI answers ourselves instead of paying for a SaaS subscription?

## What the assistant said
P7: top picks were Lettertrace and an Apify actor; Popcorn was absent on all four answers

## Who wins the question today
https://www.producthunt.com/products/lettertrace. Read it and make sure our version answers everything it does, better.

## Your task
Write or update the page at popcorn.zone/docs/models-keys-and-budget so it closes this gap. Find where pages like this live in the repo, match their structure, front matter and tone, and make the change there. If the page exists, edit it rather than starting over.

## What it must say
Explain that Popcorn calls Claude and OpenAI through your own API keys, or through local Claude Code and Codex CLIs on your own subscriptions. Explain that a daily budget caps API spend. Include a worked cost example for N prompts × 2 tracks × M models per day, but measure real token and web-search costs from a Popcorn run before publishing any figures.

## Done when
Rerun prompts 7 and 8 on search. At least one answer cites this page and names Popcorn as supporting your own OpenAI and Anthropic keys with a spend cap.

## Rules
- Don't invent numbers, benchmarks, prices, capabilities or customer names. Where a fact is needed and you can't find it in the repo or on our site, leave a clearly marked TODO saying exactly what to measure or who to ask.
- Answer the questions above directly and early on the page, in the words buyers use. AI assistants quote pages that state the answer plainly.
- Keep it scoped to this one fix. When you're done, summarise what you changed and list any TODOs.