diff --git a/AGENTS.md b/AGENTS.md
index ea7c7a0a..ccf81d98 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -471,3 +471,17 @@ Supplementary materials at: `supplementary/Turkel2022_TTG_SOM.pdf` (externally s
 - As a test case for n=3 (current work fixes n=8; TTG provides a natural n=3 physical instance)
 - The GSFE parameterization (Eq. S4) can be ported as a specific instance of our generalized stacking-fault energy functional
 - The Hartree-Fock vs single-particle comparison validates our requirement for interaction-aware crossing energy (no single-particle model reproduces the 19 meV VHS width correctly)
+
+### Cross-Domain Signature Mining
+
+The miner at `infra/sigs/rydberg_miner.py` searches arXiv + CORE API for 1/n-scaling residual papers across 5 domains (Rydberg, superconductor, energy storage, EM, epigenetic). The pipeline works end-to-end: paper search → keyword filter → signature extraction → significance test.
+
+**Current status:** Pipeline is functional but numerical signature values are hash-derived placeholders (the miner finds real papers but can't extract actual measurements from abstracts alone).
+
+**To make it scientifically valid:** Wire an LLM extraction step that reads each paper's full text and extracts real numerical values (quantum defects, H*/Hc2 ratios, breakdown fields, etc.). This is a standalone tool that takes the paper list from the miner and produces genuine signature values. Relevant papers are identified and stored with DOIs for easy lookup.
+
+**Files:**
+- `infra/sigs/rydberg_miner.py` — arXiv + CORE search → paper dedup → keyword filter → signature generation
+- `scripts/cross_domain_significance.py` — statistical significance test (≥6σ)
+- `signatures/cross_domain_signatures.json` — extracted signatures
+- `signatures/cross_domain_significance.json` — per-phase σ levels