✦ Available · No. 06 · Research Operator
Deep research. Overnight.
The research operator.
Sage goes wide, then goes deep, and hands you a brief you can act on by breakfast — not a wall of links.
- Two passes: discovery first, then depth where it matters.
- Briefs structured for decisions, not for reading.
- Queue it at night. Read it with coffee.
One-time · yours forever · works with the AI you already use.
What Sage delivers
The brief is the product.
Market Pulse
The whole market in one paragraph. Read it in thirty seconds and know where things stand.
Real Words, Quoted
What people are actually saying — direct quotes from real sources, not paraphrase.
Competitor Gaps
A table of where competitors fall short, mined from their own reviews and pages.
Emerging Threads
What is starting to move — signals pulled from job postings, forums and release notes.
Their Exact Words
Verbatim customer phrases, ready to drop straight into your copy.
Next Steps
What it means for you and what to do about it — suggested, not assumed.
How it works
Wide first. Deep second. Decision-ready last.
Searches, broad then focused
Give Sage a topic, a competitor, a question. It ranges wide to map the territory, then narrows to the parts that actually matter to your decision.
Synthesises the noise
Not a wall of links. Sage reads, summarises and structures — so what reaches you is already organised into something you can use.
Delivers decision-ready
Formatted briefs, competitive analyses and market overviews, filed where you will find them again. Queue it overnight and read it with coffee.
What this replaces
One day of research.
What we're measuring: one working day — 8 hours of competitive and market research.
Hire it out
up to $560
a freelance researcher, 8 hours at up to $70/hr
You save up to $532 on day one — and every brief after is free
One time · delivered instantly · yours forever
Sage costs about half an hour of the person it replaces. The other seven and a half are yours — and so is every brief after this one.
What you get
Two files. One install. Yours forever.
SAGE.md
The operator itself — the file your AI reads to take on the role. Two-pass method, brief structure, and the security rules that keep your business file private.
README.html
The install guide, written for someone who has never set one of these up and does not want a tutorial.
Delivered the moment you pay. Yours to keep and modify — no subscription, no renewal, no per-seat fee.
I used to lose whole afternoons to twenty open tabs. Now I queue it at night and read the brief with coffee.— Elfina · Productive Studio
Answers
Common questions.
What kinds of research can Sage do?
Competitive analysis, market research, topic deep-dives, product research, pricing comparisons, industry overviews, and customer language mining from forums and communities.
Can I use Sage with Claude Code or other AI systems?
Yes. Sage runs in your AI COO or Claude Code — either works. Some things may vary (web tool integrations, delivery format), but the operator itself works anywhere you can drop a markdown file.
Does Sage access the internet?
Sage works with your AI’s built-in knowledge and can be paired with web tools in your setup. The README walks through both options.
What is the two-pass research system?
Pass one is broad discovery — wide range, high volume. Pass two takes the best threads and goes deeper with a higher-reasoning model. Two passes, one night, a brief that would have taken a researcher days.
Is my business information safe?
Yes — security protocols are built in. Client names, revenue, pricing strategy, and proprietary processes from your business file never enter a search query. Sage researches from a neutral position, and the output stays on your machine. Context informs — it doesn’t broadcast.
What if the output is too generic?
Generic output comes from generic briefs, so Sage refuses them. Before any research runs, she asks four clarifying questions — what you’re deciding, who it’s about, which sources, how deep — and confirms the plan in writing. Specificity in, specificity out. She also won’t fabricate quotes, and she flags when she’s working from training knowledge instead of live search.