Recuris proposes agent memory without changing the model
In a publication dated August 26, 2026, Recuris describes two loops for long-horizon agents: working and experiential memories guide tasks, while a fixed meta-agent analyzes execution traces and proposes targeted repairs.
On August 26, 2026, Recuris presented an approach for long-horizon agents with two loops: working and experiential memories guide task execution; a fixed meta-agent analyzes execution traces and proposes targeted repairs to memory. The proposal aims to improve agent behavior using memory and execution feedback without changing the model's weights.
The account gives no quantitative results. To consult and verify the approach, check Recuris's original publication for details of the loops, the traces analyzed, and the proposed repairs; do not assume the description alone proves performance gains. If using AI to study or apply the material, avoid sending personal data or confidential information.