Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/eduardocornelsen/full-funnel-ai-analytics/attributiongit clone --depth 1 https://github.com/eduardocornelsen/full-funnel-ai-analyticsWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.00582 |
| Opus 5 | $0.00000 | $0.00291 |
| Sonnet 5 | $0.00000 | $0.00116 |
| Haiku 4.5 | $0.00000 | $0.00058 |
Grade A, and why
attribution scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 26 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Data sourcing (mandatory)
Read dashboards/golden_metrics.json → windowed_90d.attribution_by_channel and windowed_90d.channel_performance; copy exact values — never recalculate (CLAUDE.md §14).
- Attribution shares must sum to 100% — normalise before any pie/donut (§4).
- Label revenue figures
Linear attribution · 90d; first/last-touch columns are for model comparison only. - Freshness badge in the header:
_meta.window_start–_meta.window_end· Data as of_meta.generated_at.
Live MCP variant — only if the user appends -mcp or asks for "live" / "real-time" / "raw platform" data: query the ga4, google-ads, meta-ads MCP servers instead, passing dates from _meta.window_start / _meta.window_end, add the badge ⚡ Live MCP — may differ from golden layer, and use dashboards/js/metrics.js canonical formulas for any computed metric.
Artifact
Build a channel attribution deep-dive React artifact using Recharts.
Design: dark theme (#0d0d1a bg, #1a1a2e cards), coral #f87171, blue #60a5fa, amber #fbbf24, green #34d399.
Include these 7 sections:
- Attribution summary cards — Sessions, Conversions, Revenue, Avg CPC per channel
- Multi-channel waterfall — grouped bar chart: impressions, clicks, conversions by channel side by side
- Efficiency scatter plot — X axis: spend, Y axis: ROAS, bubble size: conversions. One bubble per campaign.
- Top/Bottom performers table — top 5 and bottom 5 campaigns by ROAS with color-coded badges
- Channel mix over time — stacked area chart showing share of spend by channel week over week
- Key insight callout — a highlighted text box summarizing the single biggest optimization opportunity
- AI Insights panel — a dark card at the bottom with a "✦ AI Insights" header containing 4–5 bullet points synthesized from attribution data. Each bullet must be specific, quantified, and actionable. Cover: (a) which channel has the best ROAS and is under-allocated relative to spend share, (b) which channel has the worst efficiency and should be cut or restructured, (c) the biggest gap between first-touch and last-touch credit (which channel drives awareness vs closes deals), (d) a CVR anomaly across channels worth investigating, (e) a concrete budget shift recommendation with estimated impact. Write in plain English as if briefing a paid media manager.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 26 lines · 0 tokens per session scan A 46347e4a0550
attribution is a command published in the GitHub repository eduardocornelsen/full-funnel-ai-analytics (21 stars, last pushed 22d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 582 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
chunkhound-release-notes
Generate release notes for a ChunkHound production release.
review-claims
Analyze patent claims for 35 USC 112(b) compliance - antecedent basis, definiteness, and claim structure.
review-epo-claims
Analyze patent claims for EPO Art. 84 EPC compliance - clarity, conciseness, support by description.
operations-agent
Deployment phase agent - build, deploy, verify, and monitor releases.
lint
Check the project for naming, security and best-practice issues.
proposal
Create a structured feature proposal in docs/11-proposals/.