Run deep research with a hard budget and verified quotes

An open-source terminal research agent enforces a spending cap, attaches a source to every claim and keeps local data local.

Difficulty
IntermediateOur rating
Time to build
Not reportedOur estimate
Monthly cost
Not reported
Revenue
Not reportedWe only show revenue the source states
Engagement
100 HN points14 comments · fetched Oct 15, 2026

What they did

Give Mole a question (and optionally a CSV); it researches within your budget, quotes sources and analyzes local files without uploading them.

How it went

The author reports 0% measured budget overshoot.

Worth knowing

This is the builder launching their own product on HN.

Builder recipe

Our own starting point, written from the source. Adapt it to your setup.

Stack

  • Mole
  • Any LLM or local model

Steps

  1. Define the question and success criteria.
  2. Let agents search, run or summarize in parallel.
  3. Verify key results yourself.
  4. Write up findings with sources.

Starter prompt

With a $1 budget, find the 5 best sources on [question], quote the exact lines that answer it, and say what remains uncertain.

Tags

Read the original

Related use cases

The Daily 5.

Five real AI agent use cases each morning, each with a recipe and a source. Plus a Sunday roundup. Free.