feat: Add LocalBoost AI template#97
feat: Add LocalBoost AI template#97viniciusnascimento-me wants to merge 1 commit intoLamatic:mainfrom
Conversation
📝 WalkthroughWalkthroughThis PR introduces a new "LocalBoost AI" template for a Lamatic workflow that scrapes local business websites using Firecrawl and generates structured lead intelligence via OpenAI's GPT-4o-mini model. The addition includes documentation, workflow configuration, input schema definitions, and template metadata files. Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes 🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (2)
templates/localboost-ai/meta.json (1)
1-9: Inconsistent indentation and empty placeholder fields.
- Line 2 uses 1-space indentation while other lines use 2-space indentation.
- The
testInput,githubUrl,documentationUrl, anddeployUrlfields are empty placeholders. Consider populating these with actual values or removing them if not needed.🔧 Suggested fix for indentation consistency
{ -"name": "LocalBoost AI – Lead Intelligence", + "name": "LocalBoost AI – Lead Intelligence", "description": "AI-powered lead analysis and outreach generation for local businesses using real website data.", "tags": "AI, Lead Generation, Sales, Automation, Local Business", "testInput": "", "githubUrl": "", "documentationUrl": "", "deployUrl": "" }templates/localboost-ai/README.md (1)
81-83: Add a language specifier to the fenced code block.The code block lacks a language identifier, which can affect syntax highlighting and accessibility. Since this is a placeholder text value, consider using
textorplaintext.📝 Suggested fix
-``` +```text your-flow-id</details> </blockquote></details> </blockquote></details> --- <details> <summary>ℹ️ Review info</summary> <details> <summary>⚙️ Run configuration</summary> **Configuration used**: Organization UI **Review profile**: CHILL **Plan**: Pro **Run ID**: `7619b948-a32a-4cce-a0e7-90400bc04ce3` </details> <details> <summary>📥 Commits</summary> Reviewing files that changed from the base of the PR and between cd0d51996202fd4981daceba6a6d9c43cd859867 and bd2d613bb87ec51f25f3ab890b955692b5a7d441. </details> <details> <summary>⛔ Files ignored due to path filters (1)</summary> * `.DS_Store` is excluded by `!**/.DS_Store` </details> <details> <summary>📒 Files selected for processing (4)</summary> * `templates/localboost-ai/README.md` * `templates/localboost-ai/config.json` * `templates/localboost-ai/inputs.json` * `templates/localboost-ai/meta.json` </details> </details> <!-- This is an auto-generated comment by CodeRabbit for review status -->
| { | ||
| "id": "187c2f4b-c23d-4545-abef-73dc897d6b7d", | ||
| "role": "user", | ||
| "content": "Business name: {{trigger.business_name}}\nWebsite: {{trigger.website}}\nInstagram: {{trigger.instagram}}\nLocation: {{trigger.location}}\nScraped website content: {{firecrawlNode_439.output.markdown}}\nAnalyze this business using the real website content above.Field instructions:- business_summary: summarize what the business does and how it positions itself- evidence: list 3 to 5 concrete signals found on the website- detected_problems: list 3 to 5 weaknesses or missing conversion/trust elements- growth_opportunities: list 3 to 5 realistic marketing opportunities- quick_wins: list 3 practical improvements that could be implemented fast- offer_angle: best service angle to pitch- p- personalized_outreach: write a short outreach message in Brazilian Portuguese that references at least one specific detail from the website (e.g. services, WhatsApp CTA, or location)- lead_score: High, Medium, or Low- reason_for_score: explain the score based on the evidence foundImportant:- Mention local positioning if present- Mention WhatsApp CTA if present- Mention service pages if present- Do not make up Instagram performance, reviews, or ad performance unless clearly supported by input\nVERY IMPORTANT:\n- You MUST include the \"evidence\" field\n- Each evidence item must reference something found in the website content\n- Do not skip this field" |
There was a problem hiding this comment.
Orphaned text p- and incorrect variable references in user prompt.
Two issues in the prompt content:
-
Orphaned text: There's a stray
"p-"before"personalized_outreach"that appears to be an editing artifact. -
Inconsistent variable references: The prompt uses
{{trigger.business_name}},{{trigger.website}}, etc., but the trigger node ID istriggerNode_1. Based on line 38 which correctly uses{{triggerNode_1.output.website}}, these should be{{triggerNode_1.output.business_name}},{{triggerNode_1.output.instagram}}, and{{triggerNode_1.output.location}}.
🐛 Suggested fix for prompt content
The prompt should reference variables consistently:
{{trigger.business_name}}→{{triggerNode_1.output.business_name}}{{trigger.website}}→{{triggerNode_1.output.website}}{{trigger.instagram}}→{{triggerNode_1.output.instagram}}{{trigger.location}}→{{triggerNode_1.output.location}}- Remove orphaned
"p- "text before"personalized_outreach"
| "generativeModelName": [ | ||
| { | ||
| "type": "generator/text", | ||
| "params": {}, | ||
| "configName": "configA", | ||
| "model_name": "gpt-4o-mini", | ||
| "credentialId": "4afd5974-b7f1-4fc8-bf04-e8841720061b", | ||
| "provider_name": "openai", | ||
| "credential_name": "Lead inte" | ||
| } | ||
| ] |
There was a problem hiding this comment.
Hardcoded credential ID and truncated credential name should be placeholders.
The generativeModelName array contains:
- A hardcoded
credentialId(4afd5974-b7f1-4fc8-bf04-e8841720061b) specific to your environment - A truncated
credential_name("Lead inte")
For a reusable template, these should be empty placeholders or clearly marked sample values that users must replace.
🔧 Suggested fix
"generativeModelName": [
{
"type": "generator/text",
"params": {},
"configName": "configA",
"model_name": "gpt-4o-mini",
- "credentialId": "4afd5974-b7f1-4fc8-bf04-e8841720061b",
+ "credentialId": "",
"provider_name": "openai",
- "credential_name": "Lead inte"
+ "credential_name": ""
}
]
What This Template Does
LocalBoost AI is a lead intelligence agent designed for local businesses.
It analyzes real website content using Firecrawl and generates structured insights, including positioning analysis, conversion gaps, growth opportunities, and personalized outreach messages.
Problem It Solves
Most outbound prospecting is generic and ineffective.
This template enables:
Providers & Prerequisites
Template Structure
Located at:
templates/localboost-ai/
Includes:
Lamatic Flow
Flow ID:
your-flow-idNotes
This is a template, not a full kit:
Use Cases
LocalBoost AI Template Addition
Summary:
Adds LocalBoost AI template—a lead intelligence agent for local businesses that analyzes website content via Firecrawl and generates structured lead insights.
Changes:
config.json — Workflow configuration with 3-step pipeline:
inputs.json — Input schemas for Firecrawl credentials and LLM model selection
meta.json — Template metadata (name, description, tags, placeholder URLs)
README.md — Documentation covering use cases, input/output schemas, control flow, and future improvements
Use Cases:
Agencies prospecting local businesses, freelancers offering marketing services, sales teams automating outbound workflows