geo-analysis — Claude Code skill: what it does, how to install
geo-analysis runs a GEO/SEO analysis for a domain through the Enception API and downloads a PDF on AI visibility, the market and competitors. Handy for client audits.
MIT 135 installs ★ 750 updated 2026-10-21
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a claude-codeWhat geo-analysis does
This skill takes a domain, starts an analysis on Enception's external API, and later fetches a PDF report. The report shows whether ChatGPT and Google AI recommend the brand, plus market research and a competitor section.
A start call needs website_url, primary_region and core_goals. When you say nothing, the skill fills in us and traffic, and you can pass up to 5 competitors by hand. A run takes roughly 20–45 minutes because the deep market-research step is slow, so check status minutes apart, not seconds. The report download needs no auth, which means anyone holding the id can open it.
Without ENCEPTION_ANALYSIS_API_KEY nothing runs, and the API is paid. The free-tier key allows 2 analyses a month, and the rate limit is 10 requests a minute. If a report shows zero visibility although ChatGPT does recommend the brand, update-brand accepts the real name and aliases and recalculates from stored answers, with no new queries.
What it helps with
- start a GEO analysis for a client's domain from one prompt
- check whether ChatGPT and Google AI recommend a brand
- download a PDF with market research and competitors
- poll a running analysis until it reaches
succeeded - fix zero visibility by giving the real brand name and aliases
Example prompts for the agent
- Run a client analysis for example.com, region us, goal traffic
- Check the analysis status and download the PDF once it finishes
- Visibility shows zero, but ChatGPT recommends the brand. Set the name to Pudu Robotics with aliases Pudu and PuduBot
How to install geo-analysis
1Skills CLI (Vercel)
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a claude-code2Manually: put the files in place
~/.claude/skills/geo-analysis/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/geo-analysis/
1Skills CLI (Vercel)
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a codex2Manually: put the files in place
.agents/skills/geo-analysis/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a cursor2Manually: put the files in place
.cursor/skills/geo-analysis/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install OpenClaudia/openclaudia-skills skills/geo-analysis2Manually: put the files in place
.github/skills/geo-analysis/SKILL.md1Skills CLI (Vercel)
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a gemini-cli2Platform CLI
gemini skills install https://github.com/OpenClaudia/openclaudia-skills --path skills/geo-analysisManually: put the files in place
.devin/skills/geo-analysis/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about geo-analysis
How do I install geo-analysis in Claude Code?
Run npx skills add OpenClaudia/openclaudia-skills --skill geo-analysis -a claude-code. The skill is the file ~/.claude/skills/geo-analysis/SKILL.md.
What does geo-analysis do?
It starts a domain analysis through the Enception API, tracks its status and downloads the PDF. The report shows whether ChatGPT and Google AI recommend the brand, along with the market and competitors.
What does geo-analysis need to work, and how long does a report take?
It needs an ENCEPTION_ANALYSIS_API_KEY, and the API is paid. The free-tier key covers 2 analyses a month. One run takes roughly 20–45 minutes.
Source text
SKILL.md · MIT
---
name: geo-analysis
description: "Run a full GEO/SEO client analysis for a domain via the Enception external API — start the analysis, poll its status, and download the PDF report covering AI visibility, market research, and competitors. Use when the user says 'run a client analysis', 'analyze <domain> for GEO', 'GEO analysis for <domain>', or 'check analysis status'."
---
# GEO Analysis
Run a full GEO/SEO analysis for a domain and get a client-ready PDF report: AI visibility (is the brand recommended by ChatGPT / Google AI), deep market research, and competitor landscape.
Thin wrapper over the Enception client-analysis API. Three operations: **start**, **status**, **report** (plus an optional brand-name correction).
**Disclosure:** contributed by Enception AI and calls Enception's **paid** external API. Docs: https://www.enception.ai/doc/api-client-analysis
## Requirements
- `ENCEPTION_ANALYSIS_API_KEY` — sent as the `x-api-key` header. Self-serve: sign up at https://www.enception.ai, then generate a free-tier key under Account → API Key (2 analyses/month free; contact the team for higher limits). Rate limit: 10 requests/minute.
## 1. Start
`POST https://www.enception.ai/api/external/client-analysis/start`
Required: `website_url`, `primary_region`, `core_goals`. Sensible defaults: `primary_region=us`, `core_goals=traffic`, `output_language=en`, `competitors=[]`. The API normalizes URLs to a bare domain and rejects IPs. Returns `{ id, message }` — **save the `id`**.
Pick `output_language` from an explicit request first, then the user's conversation language, then the domain TLD (`.cn` → `zh`; otherwise `en`).
```bash
curl -s -X POST "https://www.enception.ai/api/external/client-analysis/start" \
-H "x-api-key: $ENCEPTION_ANALYSIS_API_KEY" -H "Content-Type: application/json" \
-d '{"website_url":"https://example.com/","primary_region":"us","competitors":[],"core_goals":"traffic","output_language":"en"}'
```
Only override `primary_region` / `core_goals` / `competitors` if the user specifies. `core_goals` is free text (e.g. `traffic`, `ranking`). Competitors: `[{"name":"X","domain":"x.com"}, ...]` (max 5). For multiple domains, fire one POST each and collect the ids.
## 2. Status
`POST https://www.enception.ai/api/external/client-analysis/status` with body `{ "id": "<id>" }` and the same `x-api-key` header. Returns `{ status: 'pending' | 'failed' | 'succeeded', progress_percent?, pdf_url? }`.
A full run takes **~20–45 minutes** — the bottleneck is the deep market-research step. Poll on a long interval (minutes, not seconds). `succeeded` ⇒ ready to fetch the report. If a run fails or stalls indefinitely, report that to the user rather than inventing results.
## 3. Report
`GET https://www.enception.ai/api/external/client-analysis/report/<id>` — public, no auth; shareable by id. Works once status is `succeeded` (also serves partial runs).
```bash
curl -s -o geo-analysis-example.pdf \
"https://www.enception.ai/api/external/client-analysis/report/<id>"
```
Confirm it's a real PDF (`file geo-analysis-example.pdf`), then give the user the file and/or the public URL.
## 4. Fix "zero visibility" (brand-name correction)
`POST https://www.enception.ai/api/external/client-analysis/update-brand` with body `{ "id", "brand_name"?, "alternative_names": [...] }` and the `x-api-key` header.
When a report shows zero/low AI visibility but the brand IS recommended by ChatGPT/Google, it's almost always because the brand is only known by its fused domain label (e.g. `pudurobotics`) while AI engines say "Pudu Robotics". Supply the real name + aliases and the route recomputes visibility from the already-stored answers (no re-query — fast), then re-fetch the report.
```bash
curl -s -X POST "https://www.enception.ai/api/external/client-analysis/update-brand" \
-H "x-api-key: $ENCEPTION_ANALYSIS_API_KEY" -H "Content-Type: application/json" \
-d '{"id":"<id>","brand_name":"Pudu Robotics","alternative_names":["Pudu","PuduBot"]}'
```
Returns `{ success, mention_rate, overall_score, average_position, analysis_regenerated }`. New analyses resolve the brand name automatically; this is the manual correction path for existing reports.
© OpenClaudia · MIT · OpenClaudia/openclaudia-skills