Cryptographic Provenance Gates Mitigate Multi-Session Memory Poisoning in Autonomous Agents
Sean Maxwell — Roll Solo, Melbourne, VIC, Australia (ORCID: 0009-0004-0133-0463)
Abstract
Autonomous agent architectures deploying persistent Retrieval-Augmented Generation commit interaction history into long-term storage, introducing recursive self-poisoning when transient model confabulations emitted in Session t are retrieved in Session t+k as ground truth. We evaluate an immutable Cryptographic Provenance Gate enforced through SQLite database triggers across a pre-registered benchmark of N=100 paired multi-session probes against Naive RAG, No-Assistant Memory, and In-Prompt Soft Tagging. Under sampled adversarial decoding (T=0.7), In-Prompt Soft Tagging degrades to 40/40 (100.0%) poisoning while the Provenance Gate remains at 0/40 (0.0%), and preserves 29/30 (96.7%) benign multi-step continuity compared with 0/30 (0.0%) under assistant-turn deletion.
Citation (BibTeX)
@techreport{maxwell2026poisoning,
author = {Maxwell, Sean},
title = {Cryptographic Provenance Gates Mitigate Multi-Session Memory Poisoning in Autonomous Agents},
institution = {Roll Solo},
number = {LUCCA-TR-2026-01},
year = {2026},
doi = {10.5281/zenodo.23901452},
url = {https://doi.org/10.5281/zenodo.23901452}
}