Know when the market is building your roadmap for you.
Repoclaw reads your repository, watches AI and developer-tool releases every day, and tells you each Monday which parts of your product are still worth building.
Read-only access. Your code is never sent to a model and never used for training.
Stop or buy: sync engine. Two hosted sync services shipped offline conflict resolution this month.
Invest: template marketplace. Nobody else covers it; your usage data is the moat.
| Capability | Verdict | Score |
|---|---|---|
| Real-time sync | Commoditized | 20 |
| AI summaries | Newly enabled | 45 |
| Rich-text editor | Converging | 50 |
| Template marketplace | Defensible | 70 |
| Handwriting import | No market signal | n/a |
How it works
Repoclaw is a strategy report, not a code review. It asks one question about every part of your product: is this still worth building, given what shipped this week?
Map
Repoclaw reads your repository read-only and proposes a list of capabilities: what your product does, how much work went into each part, and what you plan next.
Confirm
You correct the list once: rename, merge, add what it missed, mark what is core. Later scans keep your list and only suggest changes.
Watch
Every day it collects releases from model labs, developer tools, open-source projects, your dependencies and the competitors you name.
Report
Every Monday you get a short email: what to invest in, what to stop building or buy, and what to watch, with a source link for every claim.
What the verdicts mean
Many providers now offer it cheaply. Building your own rarely pays.
The market is moving in. Partial versions exist; watch the gap.
Nothing in the market threatens it. Your own layer is the value.
Something new makes it cheaper or better to build. Consider using it.
Nobody visible is building this. For a niche product, good news.
Recommendation: Build (keep investing), Buy (use a product or service instead), Wrap (build a thin layer on someone else's), Watch (no change yet, keep an eye on it), Delete (only after several independent sources show it is replaced).
Score: defensibility from 0 to 100. High means hard for others to copy or replace; n/a when there is no market signal.
Confidence: how much evidence stands behind the verdict: high, medium or low.
Since last week: verdicts stay put unless new evidence arrives; when one moves, the report names the release that moved it.
Honest by design
Every verdict shows its reasoning and links to the releases behind it. A new threat starts as Watch. Repoclaw only suggests dropping something after several independent sources show it is replaced, and it never claims anything is safe to delete.
Verdicts stay stable from week to week. When one changes, the report says what changed and which new release caused it.
Openly AI
Repoclaw is built and run with AI models, and the reports are written by them. People set the rules, review the quality and answer support.
Your source code stays on our server or your own machine. Only short capability summaries go to the model, through providers whose terms exclude training. How we handle your code.
Pricing
Start with one repository for free. Upgrade when the weekly report earns its place.
Solo
$15 per month
$150 per year
for indie developers
- Up to 3 repositories
- Weekly report
- Dashboard
- Roadmap collision alerts
Team
$49 per month
$490 per year
for small teams
- Up to 15 repositories
- Weekly report and Slack or email digests
- Shared dashboard
Annual plans get two months free. Prices exclude any sales tax, which our reseller Paddle adds at checkout where required. 14-day refund on your first payment: see the refund policy.
Questions
Does it work for products that are not about AI?
Yes. For niche products it watches the competitors you name, the platforms and libraries you depend on, and the generic building blocks your features use, such as payments or translation.
Does it read my private code?
It reads it read-only, on our server or on your machine, and deletes clones after each scan. Code never goes to a model; only summaries of what each part does.
Is it a code review tool?
No. It does not comment on code quality. It tells you where to spend your building time.
Who is behind it?
Clawpy, a small independent studio in the UK. Questions go to [email protected].