QaOnFire installs statistics
By Radu Ghetu · GitHub Marketplace
QaOnFire generates AI QA reports on every pull request
QaOnFire has 4 recorded installs on GitHub Marketplace.
It ranks #1190 of 1,393 tracked GitHub Marketplace listings by installs.
PluginStats has 1 daily observations since Oct 16, 2026; the latest is from Oct 16, 2026.
About QaOnFire
AI QA reports on every pull request
Open a pull request — QaOnFire posts one comment with manual test scenarios, edge cases, setup notes, and verification scripts written for the actual diff.
A qabot.md file in your repo teaches QaOnFire your domain, so reports reference your real roles, tables, and rules — not generic boilerplate.
Free for 5 PRs/month. Works on private repos. No CI changes.
What it does
QaOnFire is a GitHub App that posts a tailored QA report on every pull request — test scenarios, edge cases, setup notes, verification scripts, and a plain-English summary for stakeholders. Unlike generic AI
code reviewers focused on syntax and style, it's purpose-built for QA.
How it works
• Install on the repos you want covered. Two clicks, no config. • Add a qabot.md to your repo root — or comment /qaonfire init and the bot opens a PR with a starter file tailored to your codebase. • Open a PR. QaOnFire reads the diff, the changed files, and your qabot.md, calls Anthropic's Claude API, and posts the report. New commits update the comment in place.
What's in every report
• Manual test scenarios — step-by-step, role-aware • Edge cases — boundaries, error paths, race conditions • Setup notes — seed data, flags, env vars • Verification scripts — SQL or shell snippets • Risks with file:line references • PM notes for stakeholders
Why qabot.md matters
A ~50-line markdown file at your repo root teaches QaOnFire your domain: roles, rules, terms, invariants. Without it, plans are generic. With it, a paranoia line like "large image uploads vs serverless 60s
timeout" becomes a concrete test scenario whenever a PR touches the upload path.
Privacy and permissions
PR content (diff + changed files + qabot.md) is sent to Anthropic's Claude API. No AI is trained on your code — not by us, not by Anthropic (per their commercial terms).
Permissions:
• Pull requests (R/W) — read metadata, post the report • Contents (R) — read the diff, files, qabot.md • Issues (R/W) — handle slash commands, open the init PR • Metadata (R) — required by GitHub
Privacy · Docs · hello@qaonfire.dev
Description from the GitHub Marketplace listing.
- Testing
- AI Assisted
Where QaOnFire sits in the listing
Each band is one page of 20. Installs on a log scale. Placement is projected from installs, so it can differ from the marketplace's own sort order. Hover a dot to see the listing.
Charts
Images: 30d · 1y · all time · 1600×900 dark
Adoption over time
7/30/90-day growth
Filled points are observed values; hollow points come from Internet Archive snapshots of the listing page. Dashed segments span periods with no observations — values between points are not interpolated.
Recent observations
Full history (1)Similar GitHub Marketplace plugins
Data source
Install counts (installations of each GitHub App across accounts and organizations) are read from public GitHub Marketplace listing pages. GitHub does not publish ratings. Category positions reflect GitHub's own popularity sort.
Listing: https://github.com/marketplace/qaonfire. First seen Oct 16, 2026; last successful collection Oct 16, 2026. See the methodology.