belumume

Cisfalcon — AI skill for Claude Code

AI community

A live pre-synthesis specificity gate for AI-designed enhancers: a frozen external measured-activity model plus a multi-agent Claude verifier.

How to install Cisfalcon

This entry records only its repository, not the path inside it, so there is no exact command to give. Open belumume/cisfalcon and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Cisfalcon does

A live pre-synthesis specificity gate for AI-designed enhancers: a frozen external measured-activity model plus a multi-agent Claude verifier. Cross-lab AUROC 0.80 on 93,435 independent designs.

Alternatives in AI

Cursor Talk To Figma MCP — TalkToFigma: MCP integration between AI Agent (Cursor, Claude Code, Codex) and Figma, allowing Agentic AI to c 7k ★
Claudish CLI — for external AI models) 679 ★
Field Lab — A field lab for thinking with AI 570 ★

README

CisFalcon

**Live tool: https://cisfalcon-lifesci.fly.dev/**

![The CisFalcon Overview tab: catch the AI-designed enhancers that will fail before the lab synthesizes them. A frozen measured-activity model gives cross-lab AUROC 0.80 against 0.50 chance on 93,435 held-out designs. Ranking safest-first and synthesizing the safer half cuts failures 41% conditioned within one target cell and one generator; pooled across a mixed batch it reads 70% (6.29% to 1.91%), but a sequence-free rule using only the stratum base rates scores about 90% pooled, so the pooled number is not evidence of per-design skill.](docs/hero_overview.png)

*A real Gosai/Tewhey design that CisFalcon flags as off-target, diagnoses, and closes the loop on: disrupt the K562 driver motifs (GATA1, TAL1, GATA2), install the HepG2 grammar (HNF4A, HNF1A, CEBPA), and the same external model moves the predicted specificity gap from failing to passing. This is an in-silico consistency check on one design, not wet-lab validation, captured live from the tool.*

In 60 seconds

**The problem.** AI now designs synthetic enhancers: short DNA meant to switch a gene on in one cell type and stay silent everywhere else, the targeting step of gene therapy. In the one published library where this has been measured end to end (93,435 Gosai/Tewhey designs, scored cross-lineage in vitro), about **1 in 16** of those designs is secretly broken and fires in the wrong cell; in-vivo cortical screens report far higher failure rates still (see Why it matters below, and `docs/failure-modes.md` for the 0.7% to 22.3% spread by generator inside this one library). You only find out after you synthesize and assay it, weeks and hundreds of dollars later.

**What it does.** CisFalcon reads a design and predicts that failure from sequence alone, before synthesis, so a lab spends its bench budget on the designs most likely to hold up.

**The number, reported the honest way.** Cross-lab **AUROC 0.80 on 93,435 designs from a different lab