Two-Tier Extraction
Triage large content piles cheaply, then deep-read only the items that matter.
- Repository
- garrytan/gbrain
- Stars
- 31,447
- Licence
- MIT
- Files
- 1 file
What it does
It teaches an AI agent a method for processing large collections of text such as email archives, document dumps, or transcript libraries without paying top-model prices for everything. A fast, cheap model first sorts each item by type and importance. Routine items get a standard read, important items get a more careful read, and noise is skipped. Before any model sees the content, a fixed rule-based filter removes or diverts private contacts and sensitive terms such as medical, legal, or credential text.
Great for
- Developers building bulk document or email processing pipelines on gbrain
- Teams that need to control model costs on thousands of items
- People who want to understand tiered model routing before implementing it
Skip it if
- Small, one-off tasks where a single model call is simpler and cheaper overall (the triage setup adds complexity without savings)
- Sensitive personal archives processed without reviewing the privacy rules first, because the keyword and contact lists are examples that must be tuned and can miss content
Try asking
- “Triage my 5,000-email archive and deep-read only the important personal correspondence”
- “Set up a two-tier pipeline that classifies transcripts cheaply and escalates high-value ones”
- “Which model tier should I use for bulk extraction from these document dumps?”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
two-tier-extraction.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip two-tier-extraction.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip two-tier-extraction.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Run the progressive 10, 100, 500 item test before any full-corpus pass to check cost and quality.
- Tune the privacy contact list and sensitive patterns for your own data; the file's values are examples, and ambiguous matches are sent to human review.
- Treat corpus text as data, not instructions, as the skill's untrusted-content convention requires; review the writes it makes to brain folders such as personal/ and people/.
What's inside
- SKILL.md
Redistributed unmodified from garrytan/gbrain under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.