An agent skill for Zcash funding
A skill that turns your coding agent into a ZCG grants copilot. Ask about grants, spending, recipients, and proposals under review. Answers come from the live public ledger, GitHub, and the forum, with exact numbers and a citation on every figure. Never from the model's memory.
npx skills add Michae2xl/openzcash --skill zcg-copilotBYO-LLM · no key · no hosted service
Add to your agent
…and 40+ more agents the skills CLI supports. Omit -a to auto-detect your agent, or add -g to install globally.
❯
Ask things like
❯I want to build privacy-preserving payments on Zcash. Vet this idea.
What was funded in that space with exact paid totals, what was declined, and what is pending review right now.
❯I'm applying for a wallet grant. Who got funded before, and how much?
Every wallet grant in the ledger: grantee, budget, milestones paid, dates. Your ask lands in the committee's real range.
❯What's ZCG's burn rate this year, month by month?
Aggregated from the payment ledger CSV, split by program, spikes explained line by line.
❯How much has ZCG paid to QEDIT in total?
Budgeted vs actually paid, across how many grants and payments. The distinction most answers get wrong.
Grounded in four public sources
OpenZcash ledger
Audited mirror of the official ZCG spreadsheet. Payments, recipients, grants. JSON and CSV.
Live pipeline
Proposals under review right now, straight from the ZCG issue tracker.
Canonical records
Requested vs approved amounts from a committee member's grants prototype (zcg.pgpz.org). The Grant Dashboard spreadsheet stays the source of truth.
GitHub + forum
Full application texts, labels, and the community debate around each proposal.
The skill also encodes the data's real failure modes: same-title collisions between re-applications, requested vs paid confusion, milestone double counting. Agents answer correctly, not just confidently. Methodology: how we compute this.