citation-monitor
Track brand mention citations across 4 LLMs (Claude, ChatGPT, Perplexity, Gemini) for a configurable list of brand keywords. Daily cron pulls write to local sqlite. Weekly markdown report shows mentio
SKILL.mdskills/seo-content-copy/citation-monitor/SKILL.md
GitHubcitation-monitor v0.1.0
Purpose
Answer the question: "Is my brand showing up when buyers ask LLMs for recommendations?"
Daily, query 4 LLMs with a fixed prompt set per tracked brand. Score each response for brand mention, list position, and citation URL presence. Store in sqlite. Render a weekly markdown report with deltas.
Architecture (v0.1.0 — local-first)
| Layer | Implementation |
|---|---|
| Inputs | tracked-brands.json (brand + keywords + niche) |
| Engines | Claude (full), OpenAI (full), Perplexity (stub), Gemini (stub) |
| Storage | sqlite at data/citations.db |
| Schedule | cron-friendly entry point scripts/run_daily.py |
| Reports | reports/<yyyy-ww>.md markdown with deltas |
Upgrade path (v0.2+): Supabase Postgres + n8n cron + multi-tenant brand list.
Schema (sqlite)
citations(
id INTEGER PRIMARY KEY AUTOINCREMENT,
brand TEXT NOT NULL,
keyword TEXT NOT NULL,
engine TEXT NOT NULL,
query TEXT NOT NULL,
response_excerpt TEXT,
mention_count INTEGER DEFAULT 0,
position INTEGER, -- 1-indexed list position; NULL if not mentioned
citation_url TEXT,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
)
Execution Flow
- Init (one-time) —
python scripts/init_db.pycreatesdata/citations.db. - Daily run —
python scripts/run_daily.pyreadstracked-brands.json, loops brand x keyword x engine x prompt-template, calls each engine, scores response, inserts row. - Weekly report —
python scripts/weekly_report.pyreads sqlite, computes week-over-week deltas, writesreports/<yyyy-ww>.md.
Triggers in conversation
When the user says any of:
- "track citations for SkynetLabs"
- "run citation monitor"
- "weekly AEO report"
/citation-monitor
Read tracked-brands.json, run scripts/run_daily.py, then offer to render the
weekly report.
Cost guardrails
- Default model IDs:
claude-haiku-4-5andgpt-4o-mini. Cheapest tier per provider. - Anthropic prompt caching enabled (5-min ephemeral) on the system prompt — every brand+keyword tuple shares the system prompt, so cache hit rate is high.
- One run = ~24 API calls (3 brands x 2 engines x 4 prompts). At Haiku + Mini pricing, approximately $0.01-0.03/day.
Constraints
- No fake-claims content. No em-dashes in any output.
- v0.1 ships Claude + OpenAI fully. Perplexity + Gemini raise
NotImplementedErrorwith a clearv0.2 — needs <KEY>message. - Python 3.11+. Pinned deps in
requirements.txt.
Files
SKILL.md— this fileREADME.md— quick startrequirements.txt— pinned deps.env.example— required env varstracked-brands.example.json— input shapescripts/init_db.py— sqlite schemascripts/run_daily.py— main loopscripts/query_engines.py— engine abstractionsscripts/weekly_report.py— markdown reportreferences/prompt-templates.md— exact promptsreferences/aeo-scoring-rubric.md— scoring mathdata/.gitkeep,reports/.gitkeep