> For the complete documentation index, see [llms.txt](https://roba-labs.gitbook.io/roba-labs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://roba-labs.gitbook.io/roba-labs/roba-the-open-robotics-development-platform/faq/quality-and-trust.md).

# Quality & Trust

**What keeps quality high?**\
Reproducible evaluation packs, community curation, usage/rating scores, stake-and-slash rules for enterprise assets, automated duplication checks, and transparent leaderboards. All challenge asset submissions require complete documentation, automated testing, and peer review.

**How does slashing work for bad actors?**\
Enterprise listings require $ROBA stakes. Fraudulent benchmarks = 100% slash; misrepresented licensing = 100% slash; non-functional assets = 50% slash. Slashed funds refund buyers and support the insurance pool.

**What if I buy a low-quality asset?**\
The insurance pool (funded by marketplace fees + slashed stakes) provides refunds. Verified fraud victims receive 100% refund. Ratings and reviews inform future buyers. If the ROBA team receives reports of exceptionally low-quality assets, they reserve the right to remove them from the platform.

**Is the team doxxed?**
\
Yes, nearly the entire team is publicly doxxed, with verifiable LinkedIn / X profiles and professional track records across robotics, Web3, and AI.