skills.rest
AMC-JTC
AMC-JTCCOMMUNITY
Explore all skills by this author →
GitHub

data-research

Extracts structured data from emails and web sources into canonical tracker pages.

Updated Aug 17, 2026
Search Tags:#structured data#data extraction#deduplication#tracker#investor updates#email parsing#yaml recipes
One-click install
npx skills add https://github.com/AMC-JTC/gbrain-1 --skill data-research-amc-jtc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: data-research
Source: https://github.com/AMC-JTC/gbrain-1/tree/main/plugin/skills/data-research
Command: npx skills add https://github.com/AMC-JTC/gbrain-1 --skill data-research-amc-jtc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning unstructured emails, attachments, and web sources into clean, deduplicated structured data is tedious and error-prone. This Skill automates the full pipeline from source search to a maintained tracker page, with safeguards against hallucinated values and duplicate entries. ## Core Features & Use Cases - Parameterized YAML Recipes: Define source queries, classification rules, extraction schemas, and tracker formats for any email-to-structured-data pipeline, with built-in recipes for investor updates, expense tracking, and company metrics. - Seven-Phase Pipeline: Search sources, classify, extract structured fields, archive raw sources, deduplicate, and update canonical tracker pages with entity backlinks. - Extraction Integrity: Saves raw sources before extraction and re-reads saved files when summarizing, preventing batch-processing hallucination bugs. - Use Case: Extract MRR, ARR, growth, burn, and runway from quarterly investor update emails, deduplicate against existing entries, and append them to a markdown tracker table with links back to each raw source. ## Quick Start Ask the agent to build a tracker for investor updates by extracting MRR and ARR from this quarter's update emails.

Frequently Asked Questions about data-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract structured data from investor update emails?▼

Use the built-in investor-updates recipe, which searches email via windowed queries, classifies messages with regex patterns, and extracts MRR, ARR, growth, burn, runway, and headcount. Results are deduplicated and appended to a markdown tracker page with links to raw sources.

How do I create a custom data extraction recipe?▼

Copy a built-in recipe from ~/.gbrain/recipes/ and edit its YAML fields: source queries, classification rules, extraction schema, tracker page path, and tracker format. The same seven-phase pipeline runs for any recipe.

How does the pipeline prevent duplicate tracker entries?▼

Before adding entries, it checks for exact matches on key fields and skips them, flags fuzzy matches (same entity and date with similar amounts) for review, and adds differing amounts with a note since they may be corrections.

Why does batch extraction sometimes produce wrong amounts?▼

LLM working memory can corrupt values after batch processing. The extraction integrity rule fixes this by saving raw sources before extraction and re-reading saved files when summarizing, never trusting in-memory values.

Can it process PDF attachments and documents?▼

Yes. Attachments are handled through PDF extraction and HTML stripping, then archived via file_upload. Large files get .redirect.yaml pointers in storage, and every tracker entry links back to its raw source.

Related Skills

Explore More Data & Analytics Skills→
ComposioHQComposioHQ
74.2k

scrapfly-automation

Automate web scraping workflows and data extraction tasks using the Scrapfly API.

Official
Intermediate
Timmy6942025Timmy6942025
3

portfolio-monitoring

Extract and analyze portfolio company financial metrics against budget and plan.

Community
Advanced
ngochuy13ngochuy13

signal-detector

Capture original thinking and entity mentions from inbound messages into knowledge pages.

Community
Advanced
jasrajtulsijasrajtulsi

polars

Convert pandas pipelines into optimized Polars DataFrame operations.

Community
Advanced