Foglift CLI

Why a CLI?

Most AEO and GEO tools live entirely inside a dashboard. That’s fine for the first scan but it doesn’t scale: every iteration on a pricing page or a comparison table means a screenshot, a paste into Slack, a tab switch, a re-scan, another paste. The dashboard hop is the slow path.

A CLI collapses the loop. foglift scan takes one URL and returns a structured score; --threshold=85 turns the score into a build gate; --json pipes into jq, gh, or any shell tool you already use. The same workflow that makes Lighthouse CI a default in performance work — drop a CLI invocation into your pipeline, fail the build when scores drop — applies one-to-one to AEO regressions.

The cost of not having a CLI is concrete. AEO scores drift the same way performance scores drift: someone adds a third-party script, a refactor strips a heading hierarchy, a CMS migration breaks an FAQ schema block. Without a CLI, the regression is invisible until the next dashboard check, which is usually after the deploy. With a CLI in CI, the merge is blocked before the regression ships.

The twelve commands

Every command targets one Foglift API surface. The first three work without an API key; the remaining nine require FOGLIFT_API_KEY.

Setup

1. Install (or skip and use npx)

# Global install
npm install -g foglift-scan

# Or run any command on demand without installing
npx foglift-scan https://example.com

2. Run your first scan (no auth required)

foglift scan https://example.com

# Machine-readable output
foglift scan https://example.com --json

# Fail the shell when overall score < 85
foglift scan https://example.com --threshold=85

3. Generate an API key for the authenticated commands

Sign up at foglift.io, then Dashboard → Settings → API Keys. Free plan is fine; the key is prefixed sk_fog_. Export it and the CLI picks it up automatically:

export FOGLIFT_API_KEY="sk_fog_..."

# Now batch, AI Visibility, history, prompts, models, sentiment all work
foglift scan batch https://foglift.io https://stripe.com https://ramp.com
foglift scan ai-check --prompt "best AEO tool" --models chatgpt,perplexity --domain foglift.io
foglift scan results --days 30 --model chatgpt --json

Three terminal workflows worth automating

CI/CD: gate the deploy on AEO score

Drop the CLI into a pipeline step after the preview deploy. The job fails when scores regress; the merge is blocked until the regression is fixed.

# .github/workflows/aeo-gate.yml
name: AEO gate
on: pull_request
jobs:
  aeo:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Wait for preview deploy
        run: ./scripts/await-preview.sh
      - name: Foglift AEO scan
        env:
          FOGLIFT_API_KEY: ${{ secrets.FOGLIFT_API_KEY }}
        run: npx foglift-scan "${PREVIEW_URL}" --threshold=85
      - name: Archive scan JSON
        if: always()
        run: npx foglift-scan "${PREVIEW_URL}" --json > scan.json
      - uses: actions/upload-artifact@v4
        with:
          name: foglift-scan
          path: scan.json

The first invocation gates; the second archives. Two CLI calls, no wrapper script. The same shape works in GitLab CI, CircleCI, and Buildkite — the CLI relies only on standard Unix exit codes.

Headless competitor sweep

Run a batch of competitors in one command and pipe to jq for a leaderboard. Useful for weekly category reports and pre-launch positioning checks.

foglift scan batch \
  https://foglift.io \
  https://tryprofound.com \
  https://peec.ai \
  https://otterly.ai \
  --json \
  | jq '[.results[] | {url, aeo: .aeoScore, geo: .geoScore}] | sort_by(.aeo) | reverse'

AI Visibility delta after a content change

After shipping a fix, re-check whether ChatGPT or Perplexity actually started citing the page. The ai-check command queries the engines live; results pulls the trend over a window.

# Live check on a single prompt
foglift scan ai-check \
  --prompt "best tool to track AI search visibility" \
  --models chatgpt,perplexity,claude,gemini \
  --domain foglift.io

# Trend over the last 30 days for a specific model
foglift scan results --days 30 --model chatgpt --json \
  | jq '[.[] | {date, citationRate}] | sort_by(.date)'

The CI/CD pattern: gate, capture, comment

The CLI’s value is highest when it lives in a pipeline. The minimal-but-complete recipe has three steps that each map to a single CLI invocation:

Why this pattern compounds: tight feedback loops are the dominant productivity factor in iterative software work — documented across decades of empirical research, from Knight & Leveson on rapid prototyping (1990) through Lighthouse CI, which has shipped the same gate-capture-comment shape since 2019. Performance regressed quietly until Lighthouse CI made it a required check; AEO regresses the same way until a CLI gate makes it visible. The tooling is the difference; the loop is the same.

Foglift CLI vs. other AI search optimization tools

CLI parity across the AI search optimization category as of April 2026:

See the broader AI search tool landscape in our Best AI Search Tools with MCP Integration 2026 and Best GEO Tools for Developers breakdowns.

Sources & data freshness

All competitor parity claims verified against vendor public pages or public npm registries on 2026-05-02. CLI version pinned to the latest published release.

1. Foglift. foglift-scan on npm — package, version 1.0.1, license, README. Accessed 2026-05-02.
2. Google Chrome. Lighthouse CI — gate-capture-comment pattern reference. Accessed 2026-05-02.
3. Knight, J. C., & Leveson, N. G. (1990). An experimental evaluation of the assumption of independence in multiversion programming — feedback-loop research foundation.
4. Aggarwal et al. Agentic refactoring loops at KDD 2024 — empirical productivity gains from collapsing manual cycles into agent loops.
5. npm registry searches for foglift-scan, profound-cli, peec-cli, otterly-cli, athena-cli, semrush-ai-cli, ahrefs-brand-radar-cli — none of the non-Foglift packages exist on the public registry as of 2026-05-02.
6. Foglift. /pricing — plan tier and free-key availability self-verification.