2. Summary
Rivet is a decentralized developer platform that backs up, timestamps, and restores GitHub history using CKB, IPFS, and Arweave. V1 is live on testnet with CCC wallet onboarding, a functional backup engine, 1 click restore, and public developer profiles. The testnet currently runs on free tier hosting.
This proposal requests a grant of $5,000 for Phase 1 of development, covering CoTA frictionless onboarding, the Sovereign Passport (RGB++ Spore DOB), and the production release of the Rivet CLI restore tooling. A follow up proposal will be submitted later for Phase 2, which covers CKB Mainnet deployment, migration to a dedicated domain and VPS hosting, production infrastructure, and audit preparation.
Grant Amount Requested (Phase 1): $5,000
ETA to Completion of Phase 1: 8 weeks from grant approval
CKB Wallet or Funding Address: ckb1qyqvx63t4e0zjk56lldkre9w0225dmuthneq7texcs
Live Demo: https://rivet-web-ten.vercel.app
3. Project Introduction
What problem are you solving?
Developers build their careers, reputations, and open source contributions on platforms they do not control. GitHub can suspend an account overnight with no real appeal process, erasing years of work. If a developer loses access to their account for any reason (suspension, hack, forgotten password, or loss of an old email), they also lose their contribution graph and their proof of authorship. Git commit timestamps are trivially easy to forge, and there is no native cryptographic proof that a specific developer wrote a specific piece of code at a specific time.
What ecosystem or user need does this address?
There are millions of developers worldwide who depend on GitHub as the primary record of their professional work. Every one of them is exposed to the risk of losing that record without warning. Rivet gives developers a decentralized insurance policy. It mirrors their entire repository history to IPFS and Arweave, anchors a cryptographic proof of that history to a CKB Cell, and issues a Sovereign Passport (RGB++ Spore DOB) that proves their developer identity on chain. If anything happens to their account, they can restore everything to a new account with the original history intact.
Why now?
Centralized platforms are becoming less reliable and more censorious. At the same time, CKB’s tooling has matured to the point where this product can be built properly. CCC provides seamless wallet authentication, RGB++ enables Bitcoin bound asset ownership, CoTA enables account abstraction for frictionless onboarding, and Spore DOBs allow for flexible on chain identity. This is the right moment to give developers a real, practical, decentralized alternative for preserving their work.
4. Team and Roles
Name / Handle: Jedi Dtechmaker (jedi_dtechmaker)
Role: Lead Developer, Product Designer, and Project Manager
Experience / Relevant Background:
I am a full stack and Web3 developer with hands on experience across TypeScript, Solidity, JavaScript, Node.js, Go, Rust, Python, and PHP. I have worked extensively with the CKB ecosystem, including CCC for wallet integration and transaction signing, RGB++ for Bitcoin bound assets, and Fiber Network for Layer 2 payments. I am building Rivet because I have personally experienced the frustration of platform dependence and want to give developers a permanent, cryptographically verifiable home for their work.
Links:
GitHub: jedi-dtechmaker (Jedi Dtechmaker) · GitHub
Twitter/X: @jedi_dtechmaker
5. Current Status
Research completed:
Studied existing backup tools (Octarchive, GitPreserver, Vector migrate) and identified their limitations. Researched CKB Cell capacity costs, RGB++ Spore DOB mechanics, CoTA account abstraction, and CCC integration patterns. Defined the two mode restore architecture (Exact Restore versus Attribution Restore) after community feedback.
Prototype and MVP status:
V1 is live on testnet and functional. The web platform is deployed on Vercel and uses GitHub Actions as the backup worker. The CKB anchoring logic is running on testnet with a public verifier page.
What is live and working in V1 (testnet):
GitHub OAuth sign in with one click repo import.
Wallet connect via CCC (JoyID, MetaMask, UniSat). Optional, not required for backups.
Full mirror backup: every branch, tag and commit, dispatched to a GitHub Actions worker.
Merkle root computed over the entire commit history and anchored in a CKB testnet cell, verifiable on chain.
Git bundle pinned to IPFS via Pinata.
Proof panel on every repo with a direct CKB Explorer link.
1 Click Restore on the web: rehydrate a repo with full commit attribution intact.
Auto sync: every git push triggers a fresh backup via GitHub webhook.
Public verify page to look up any proof by CKB tx hash, IPFS CID, or GitHub repo ID.
Public developer profiles, file tree browser, commit history, and README badge generator.
Fully responsive UI with dark mode.
Privacy note for testnet:
Because Rivet is currently serverless, backups run through a public GitHub Actions queue. Private repo names and their IPFS CIDs appear in public logs. AES 256 encryption for private repos is in development. For now, users are advised to back up only public repositories. This is a temporary testnet constraint and will not exist on Mainnet.
Hosting note for testnet:
The platform is currently deployed on Vercel free tier and depends on GitHub Actions for the backup worker. Phase 2 will migrate Rivet to its own dedicated domain and a production VPS, removing the public GitHub Actions dependency entirely and enabling private repository support.
What Phase 1 of this grant will fund:
CoTA frictionless onboarding, the Sovereign Passport, and the production release of the Rivet CLI. A separate Phase 2 proposal will cover CKB Mainnet deployment, dedicated domain and VPS hosting, and production infrastructure.
Community discussion:
I posted the project on CKBuilders Review group and Talknervos and received detailed technical feedback. The key points raised were: separating Exact Restore from Attribution Restore, rewording the timestamp claim to prove repository state existed by a certain block height rather than proving git timestamps themselves, and using CoTA for frictionless onboarding so users never buy CKB. All of these have been incorporated into the design.
6. Application Design
6.1 Functional Overview
Rivet has two interfaces that share the same core engine: a web dashboard for easy onboarding and profile management, and a CLI for developers who want to automate, script, or run restores directly from their terminal.
Web dashboard flow:
A developer signs in with GitHub OAuth. No wallet is required to start. On first backup, a JoyID burner wallet is created silently in the background using WebAuthn and passkeys. Rivet subsidizes the approximately 150 CKB needed to create the user CoTA cell. The developer’s identity and backup proof are stored in a single CoTA Sparse Merkle Tree, compressed into one 32 byte hash on chain.
Rivet then clones the repositories using git clone --mirror, generates a Merkle Tree from the commit hashes, pins the bundle to IPFS via Pinata, and anchors the Merkle root to a CKB Cell. The developer receives a Sovereign Passport (RGB++ Spore DOB) that represents their on chain developer identity.
CLI flow:
The Rivet CLI is a cross platform binary for Linux, macOS, and Windows. Developers can run it locally or inside a CI/CD pipeline.
rivet backup <repo>triggers a mirror backup and prints the CKB tx hash and IPFS CID.rivet status <repo>shows the latest anchored proof and block height.rivet verify <tx_hash>independently verifies a proof against CKB and IPFS.rivet restore <repo> --mode exactor--mode attributionperforms a full local restore, usinggit-filter-repofor attribution mode and anchoring a signed hash mapping proof on CKB.rivet watch <dir>monitors a local repository and triggers a backup on every push.
If the developer ever loses access to their GitHub account, they run the Rivet CLI restore command. They choose between Exact Restore (which preserves original commit hashes and the Merkle proof) or Attribution Restore (which remaps old emails to a new email so the contribution graph lights up). Their repositories are pushed to the new account with full history intact, entirely from the terminal.
On chain: CKB Cell storing Merkle roots, RGB++ Spore DOB for identity, CoTA Sparse Merkle Tree for compressed identity storage.
Off chain: Git bundles, IPFS and Arweave storage, PostgreSQL metadata, Next.js frontend, GitHub Actions worker, Rivet CLI.
6.2 Architecture and Design
Smart Contracts and Scripts:
CKB Timestamp Cell script for anchoring Merkle roots.
RGB++ Spore DOB script for the Sovereign Passport.
CoTA Sparse Merkle Tree for compressed identity storage.
Key CKB Features Used:
CCC (@ckb-ccc/core and @ckb-ccc/connector-react) for wallet auth and transaction signing.
CKB Cell model for immutable, decentralized storage of Merkle proofs.
RGB++ for Bitcoin bound asset ownership of the Sovereign Passport.
CoTA for account abstraction and frictionless onboarding.
Spore Protocol for flexible on chain developer identity.
External Dependencies:
IPFS (Pinata) for hot storage.
Arweave for permanent cold storage.
GitHub Actions for backup worker (testnet only, to be replaced in Phase 2).
GitHub OAuth for repository access.
CLI Stack:
Rust or Node.js single binary distribution. git2 or isomorphic-git for local git operations. @ckb-ccc/core for CKB transaction signing.
All core scripts will be open sourced. The full codebase will be publicly available on GitHub.
6.3 Design Rationale
Trade offs considered:
Storing every commit hash on chain would be prohibitively expensive. I use Merkle trees to anchor a single root that represents thousands of commits. This keeps CKB capacity costs low while preserving cryptographic verifiability. For restore, I separate Exact Restore from Attribution Restore because rewriting Git history changes commit hashes, which would break the link to the anchored Merkle root. Exact Restore preserves cryptographic integrity. Attribution Restore offers a signed mapping proof anchored on CKB so the proof chain remains verifiable even after email remapping. For onboarding, I use CoTA because it compresses an entire user dataset into a single 32 byte hash on chain, which makes it viable to subsidize user onboarding and give Web2 developers a seamless experience. The CLI exists because developers need a local, scriptable tool, not just a web UI.
Security considerations:
The Merkle root anchored on CKB proves that a specific repository state existed by a certain CKB block height. It does not prove that the internal git commit timestamps are truthful. I am transparent about this distinction in all UI messaging, CLI output, and documentation.
Alignment with CKB philosophy:
CKB is designed for decentralized state verification and true ownership. Rivet uses CKB exactly as intended: as a permanent, uncensorable notary for developer owned data.
6.4 Fee Model and Sustainability
Rivet will be free for individual developers for basic backup and restore. A premium tier will offer larger storage quotas, private repository backups, priority IPFS pinning, and team features. A small fee will be charged for Arweave permanent storage. Sovereign Passport minting will carry a nominal CKB fee.
Additional revenue streams include B2B API access for credential verification by recruiters and HR platforms, protocol level fees on on chain bounty payments using a Sovereign Passport as escrow, and cryptoeconomic storage staking where node operators stake CKB to guarantee long term repository pinning.
As the user base grows, the platform can sustain itself through these premium features without compromising the free core.
7. Key Benefits for CKB
Network growth: Every backup anchors a transaction on CKB, driving real on chain activity from developers who would not otherwise interact with the network.
Addressing a need: Provides a concrete, practical use case for CKB that solves a real problem for developers worldwide.
Developer tooling: Demonstrates the power of CCC, RGB++, CoTA, and Spore in a production grade application, and ships a cross platform CLI that any developer can install and use immediately.
Community engagement: Onboards Web2 developers into the CKB ecosystem through a tool they actually need.
Interoperability: Uses RGB++ to bind developer identity to Bitcoin, showcasing CKB’s unique position as a Bitcoin Layer 2.
8. Detailed Deliverables and Milestones (Phase 1)
This proposal funds Phase 1 only. Phase 2 will be proposed separately.
| Milestone | Deliverable(s) | Timeline | Budget |
|---|---|---|---|
| Milestone 1: CoTA Onboarding and Sovereign Passport | CoTA account abstraction integration, JoyID burner wallet creation flow, Rivet subsidized CoTA cell creation, RGB++ Spore DOB minting, public profile integration. Acceptance: end to end onboarding and passport minting live on testnet. | Weeks 1 to 4 | $2,500 |
| Milestone 2: CLI Restore Tooling and Hardening | Cross platform CLI for Linux, macOS, and Windows. Exact and Attribution restore modes, git-filter-repo integration, rivet backup, rivet status, rivet verify, and rivet restore commands. Robust RPC error handling, Auto Sync reliability. Acceptance: CLI published and restore flow verifiable on testnet. |
Weeks 5 to 8 | $2,500 |
Total Budget for Phase 1: $5,000
9. Budget Breakdown (Phase 1)
| Deliverable | Budget |
|---|---|
| Milestone 1: CoTA Onboarding and Sovereign Passport | $2,500 |
| Milestone 2: CLI Restore Tooling and Hardening | $2,500 |
| Total | $5,000 |
What each milestone covers:
Milestone 1 ($2,500): CoTA Onboarding and Sovereign Passport
- CoTA account abstraction integration
- JoyID burner wallet creation flow using WebAuthn and passkeys
- Rivet subsidized CoTA cell creation (approx. 150 CKB per user)
- RGB++ Spore DOB minting for the Sovereign Passport
- Public profile integration to display the passport
- End to end testing on testnet
Milestone 2 ($2,500): CLI Restore Tooling and Hardening
- Cross platform CLI binary for Linux, macOS, and Windows
- Exact and Attribution restore modes
git-filter-repointegration for email remappingrivet backup,rivet status,rivet verify, andrivet restorecommands- Robust RPC error handling and retry logic
- Auto Sync reliability improvements
- CLI documentation and usage guide
Infrastructure costs (premium IPFS pinning, production VPS) and audit costs are excluded. They belong to Phase 2.
Payment Schedule:
- Upfront (50%): $2,500 released upon proposal approval.
- Upon Completion (50%): $2,500 released after both Phase 1 milestones are delivered and verifiable on testnet.
10. Out of Scope and Future Funding Needs
This proposal covers Phase 1 only. The following items will be addressed in a separate Phase 2 funding request once Phase 1 milestones are delivered:
- CKB Mainnet deployment: Migrating the anchor layer from testnet to Mainnet.
- Dedicated domain and VPS hosting: Moving Rivet off Vercel free tier onto its own domain and a production VPS. This removes the dependency on public GitHub Actions for the backup worker and enables private repository support with AES 256 encryption before IPFS upload.
- Production infrastructure: Premium IPFS pinning, production VPS for cron jobs and the verifier indexer.
- Security review: Formal third party audit of the CKB scripts and CoTA integration.
- Long term maintenance: Ongoing IPFS pinning costs and infrastructure scaling.
The Phase 2 proposal will be submitted only after Phase 1 deliverables are complete and verifiable on testnet.
11. Risk and Mitigation
| Risk | Mitigation |
|---|---|
| Technical complexity of Merkle proof and restore flow | Phased approach with working V1 already live on testnet, open source code, and community feedback at every milestone |
| Solo developer resource constraints | Clear milestones, realistic scope for Phase 1, and active community support |
| Dependencies on external protocols (IPFS, Arweave) | Redundant storage, fallback to local downloads, and monitoring |
| Community adoption challenges | Free tier, CoTA frictionless onboarding so no CKB purchase is required, cross platform CLI for power users, README badge, GitHub Action, and a public demo showing backup, delete, restore, and verify |
| CoTA subsidy economics | Monitor user acquisition cost per onboarded developer. Premium tier revenue must cover subsidy before scaling |
| Phase 2 funding not approved | Phase 1 deliverables are self contained and usable on testnet. If Phase 2 is delayed, the project still functions for testing and community use |
12. Closing and Call to Action
Rivet is a practical, developer first tool that brings real utility to CKB. This proposal funds Phase 1, which covers the CoTA onboarding and Sovereign Passport that make the product accessible to non crypto developers, plus the production release of the Rivet CLI that makes the core promise of backup, verify, and restore real for developers who work in the terminal. A Phase 2 proposal will follow for Mainnet deployment, dedicated domain and VPS hosting, and audit preparation. I appreciate your consideration and welcome any feedback or questions from the community.
13. Supporting Links
GitHub Repo: GitHub - jedi-dtechmaker/Rivet · GitHub
Live Demo: https://rivet-web-ten.vercel.app
Architecture Doc: Rivet/docs/ARCHITECTURE.md at main · jedi-dtechmaker/Rivet · GitHub
DAO Proposal Sync (Discussion Phase)