← Back to Skills Marketplace
danyangliu-sandwichlab

Funnel Helper

by danyangliu · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
315
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install funnel-ads-helper
Description
Diagnose and optimize full conversion funnels for paid traffic from Meta (Facebook/Instagram), Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, and Shopify A...
README (SKILL.md)

Funnel Helper

Purpose

Core mission:

  • Analyze conversion funnel drop-off by stage.
  • Identify bottlenecks from ad click to checkout or lead submit.
  • Recommend stage-specific optimization actions.
  • Define funnel experiment roadmap and expected impact.

When To Trigger

Use this skill when the user asks for:

  • conversion funnel diagnosis
  • CVR optimization planning
  • landing page and checkout improvement sequence
  • funnel experiment design tied to ROAS/CPA goals

High-signal keywords:

  • conversion, funnel, checkout, cvr
  • cpa, roas, traffic, landing page
  • campaign, optimize, retarget

Input Contract

Required:

  • funnel_stage_metrics
  • traffic_source_breakdown
  • conversion_goal
  • observation_window

Optional:

  • session_replay_notes
  • form_or_checkout_logs
  • segment_breakdowns
  • experiment_history

Output Contract

  1. Funnel Stage Health Scorecard
  2. Bottleneck Priority Ranking
  3. Optimization Actions by Stage
  4. Experiment Roadmap with KPI impact
  5. Monitoring and Iteration Rules

Workflow

  1. Normalize funnel definitions and stage metrics.
  2. Rank drop-off severity and opportunity size.
  3. Map root causes (message mismatch, UX friction, trust gap, etc.).
  4. Recommend stage-specific actions and experiments.
  5. Define monitoring thresholds and iteration cadence.

Decision Rules

  • If top-funnel CTR is strong but CVR is weak, prioritize LP and checkout fixes.
  • If add-to-cart is strong but purchase is weak, prioritize trust/payment friction fixes.
  • If retargeting conversion is low, review audience freshness and offer relevance.
  • If funnel data is sparse, run diagnostic experiments before major redesign.

Platform Notes

Primary scope:

  • Meta (Facebook/Instagram), Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, Shopify Ads

Platform behavior guidance:

  • Keep funnel interpretation tied to traffic intent by channel.
  • Distinguish ad-side and on-site bottlenecks before action.

Constraints And Guardrails

  • Do not infer funnel causes without stage-level evidence.
  • Keep test queue prioritized by expected impact and effort.
  • Avoid simultaneous high-impact changes that break attribution clarity.

Failure Handling And Escalation

  • If stage definitions are inconsistent, output a canonical funnel mapping first.
  • If missing checkout data blocks diagnosis, request minimum event payload.
  • If conversion drops sharply during active changes, trigger rollback review.

Code Examples

Funnel Health Schema

stages:
  - impression_to_click
  - click_to_viewcontent
  - viewcontent_to_addtocart
  - addtocart_to_checkout
  - checkout_to_purchase
primary_metric: stage_cvr

Bottleneck Prioritization Rule

impact_score = dropoff_pct * traffic_volume * margin_weight
sort_by: impact_score_desc

Examples

Example 1: CVR collapse

Input:

  • Click volume stable, purchases down

Output focus:

  • stage bottleneck map
  • immediate fixes
  • monitor plan

Example 2: Checkout friction

Input:

  • Add-to-cart high, checkout completion low

Output focus:

  • checkout friction hypotheses
  • test sequence
  • expected lift range

Example 3: Funnel rebuild plan

Input:

  • Multi-channel traffic with inconsistent landing paths

Output focus:

  • canonical funnel design
  • stage KPI definitions
  • experiment roadmap

Quality Checklist

  • Required sections are complete and non-empty
  • Trigger keywords include at least 3 registry terms
  • Input and output contracts are operationally testable
  • Workflow and decision rules are capability-specific
  • Platform references are explicit and concrete
  • At least 3 practical examples are included
Usage Guidance
This skill appears coherent and low-risk because it only operates on data you supply. Before using it, avoid sharing raw logs or session replays that contain PII, auth tokens, or customer identifiers — prefer aggregated or anonymized metrics. If you want automated fetching from ad/analytics platforms, require an explicit connector that limits scopes (avoid pasting API keys into free-text prompts). Validate any recommended experiments in a staging environment and confirm the agent does not attempt to reach external endpoints or ask you to paste credentials into chat.
Capability Analysis
Type: OpenClaw Skill Name: funnel-ads-helper Version: 1.0.0 The 'funnel-ads-helper' skill is a standard marketing analysis tool designed to help users optimize advertising conversion funnels across platforms like Meta, Google, and TikTok. The bundle consists entirely of descriptive markdown and metadata (SKILL.md, metadata.json) with no executable code, shell commands, or network activity. There are no indicators of data exfiltration, prompt injection, or malicious intent.
Capability Assessment
Purpose & Capability
Name/description (funnel diagnosis and optimization for paid channels) aligns with the inputs and outputs declared in SKILL.md; no unexpected credentials, binaries, or platform installs are requested.
Instruction Scope
Runtime instructions limited to analyzing user-provided funnel metrics, producing scorecards, bottleneck maps, and experiment plans. It does not instruct accessing system files, environment variables, or external endpoints. Note: optional inputs (session replay notes, logs) may contain sensitive user data but are explicitly user-supplied.
Install Mechanism
No install spec and no code files (instruction-only). This minimizes on-disk execution and is the lowest-risk install model.
Credentials
No environment variables, credentials, or config paths are required. Requested inputs are application metrics and logs, which are proportionate to a funnel diagnosis skill.
Persistence & Privilege
Flags show no always:true, no install-time persistence, and the skill does not request system-wide changes or elevated privileges.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install funnel-ads-helper
  3. After installation, invoke the skill by name or use /funnel-ads-helper
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of funnel-ads-helper – a skill for diagnosing and optimizing paid traffic conversion funnels. - Analyzes funnel drop-off by stage across Meta, Google, TikTok, YouTube, Amazon, and Shopify Ads. - Identifies conversion bottlenecks and prioritizes optimization actions. - Provides a structured experiment roadmap tied to campaign goals (ROAS/CPA). - Includes clear input/output contract and practical guided examples. - Built-in decision rules, platform-specific guidance, and failure handling protocols.
Metadata
Slug funnel-ads-helper
Version 1.0.0
License
All-time Installs 1
Active Installs 1
Total Versions 1
Frequently Asked Questions

What is Funnel Helper?

Diagnose and optimize full conversion funnels for paid traffic from Meta (Facebook/Instagram), Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, and Shopify A... It is an AI Agent Skill for Claude Code / OpenClaw, with 315 downloads so far.

How do I install Funnel Helper?

Run "/install funnel-ads-helper" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Funnel Helper free?

Yes, Funnel Helper is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Funnel Helper support?

Funnel Helper is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Funnel Helper?

It is built and maintained by danyangliu (@danyangliu-sandwichlab); the current version is v1.0.0.

💬 Comments