Bitcoin Kernel MCP
Third PlaceDon't let your AI agent reason about Bitcoin consensus. Let it ask the consensus engine directly: an MCP server exposing libbitcoinkernel for hallucination-free script and block validation.
TEAM · 1
§01 · EVENT BADGES
6 editions attended.
§02 · HACKATHON PROJECTS
Built at the hackathon.
TEAM · 1
§03 · TALKS & PANELS
On stage.
talk
· Toronto 2026
CISA
Cross-input signature aggregation (CISA) was part of the original promise for adopting Schnorr signatures in Bitcoin, but it did not end up as a part of the Taproot soft fork and was never even formalized until now. With CISA, the inputs of a transaction share signature data instead of each carrying its own signature. That makes transactions with many inputs smaller and cheaper, and it makes CoinJoins cheaper per participant the larger they get. This talk will present the three BIPs that now specify a transaction-wide CISA proposal: BIP 458 for half-aggregation, which is non-interactive, BIP 459 for full-aggregation, which compresses all signatures into a constant size but requires an interactive signing session, and a new consensus BIP that adds both to Bitcoin with a new witness version. Each input chooses its own aggregation mode, and no signature can be aggregated without the consent of its signer.
panel
· Nairobi 2026
Working on Bitcoin
We invite contributors to bitcoin-core to tell us about their journey working in public.
talk
· Nairobi 2026
Be Robin: A path into Bitcoin Open Source
Everyone wants to be Batman, a superhero with their own marquee project, attracting constant review, funding, and recognition. For newcomers to Bitcoin open source, chasing that ambition straight out of the gate is often a trap. There's a much better opening move: being Robin. Find someone whose project already has momentum, and support them. Why this works and how to pull it off is what this talk is about, including a lot of practical advice.
talk
· Riga 2025
CISA
- Pros and Cons of a CISA softfork (what's in the HRF Report) - Current state of proposal and implementation FWIW, I could speak longer than 20min but I like to keep it short :) But if there is still room in the schedule let me know and I can extend the content.
whois/fjahr
records: public · permanent