<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2026-prism">
    <titleInfo>
        <title>Prism: Persistent Faceted File Memory for Cost-Efficient Coding Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Ziyun</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yihao</namePart>
        <namePart type="family">Liu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuzhe</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Huaqian</namePart>
        <namePart type="family">Cai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>Coding agents working on sequential maintenance tasks re-read the same files across related tasks, wasting more than a third of their input tokens on redundant context. The natural fix, caching file summaries, fails to reliably reduce cost: tool-use traces are consistent with agents over-relying on generic descriptions, skipping re-reads, and offsetting token savings with less effective downstream decisions, leaving net cost roughly unchanged. We introduce Prism, a content-hash-keyed file memory that augments summaries with bounded facets: identifier-grounded facts distilled from the agent’s own prior reasoning. On 12 SWE-Chain dependency-upgrade chains, Prism reduces cost by 24.9% on GPT-5 and 9.9% on GPT-5.4-mini, while maintaining comparable aggregate Build+Fix F1. Against no memory, a summary-only ablation yields only a nominal 2.4% cost decrease that falls within run-to-run variability, while a static Aider-style repo-map baseline costs substantially more at 13% over no memory. Two controlled ablations isolate the causal contributions: wiping the persistent store before every task raises cost by 37% over no memory, and replacing facets with a 3× larger unstructured summary raises it by 41%. Code: https://github.com/WhiteFlamesCN/Prism-Opencode.</abstract>
    <identifier type="citekey">zhang-etal-2026-prism</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.1301/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>27518</start>
            <end>27531</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Prism: Persistent Faceted File Memory for Cost-Efficient Coding Agents
%A Zhang, Ziyun
%A Liu, Yihao
%A Zhang, Yuzhe
%A Cai, Huaqian
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhang-etal-2026-prism
%X Coding agents working on sequential maintenance tasks re-read the same files across related tasks, wasting more than a third of their input tokens on redundant context. The natural fix, caching file summaries, fails to reliably reduce cost: tool-use traces are consistent with agents over-relying on generic descriptions, skipping re-reads, and offsetting token savings with less effective downstream decisions, leaving net cost roughly unchanged. We introduce Prism, a content-hash-keyed file memory that augments summaries with bounded facets: identifier-grounded facts distilled from the agent’s own prior reasoning. On 12 SWE-Chain dependency-upgrade chains, Prism reduces cost by 24.9% on GPT-5 and 9.9% on GPT-5.4-mini, while maintaining comparable aggregate Build+Fix F1. Against no memory, a summary-only ablation yields only a nominal 2.4% cost decrease that falls within run-to-run variability, while a static Aider-style repo-map baseline costs substantially more at 13% over no memory. Two controlled ablations isolate the causal contributions: wiping the persistent store before every task raises cost by 37% over no memory, and replacing facets with a 3× larger unstructured summary raises it by 41%. Code: https://github.com/WhiteFlamesCN/Prism-Opencode.
%U https://aclanthology.org/2026.findings-emnlp.1301/
%P 27518-27531

Markdown (Informal)

[Prism: Persistent Faceted File Memory for Cost-Efficient Coding Agents](https://aclanthology.org/2026.findings-emnlp.1301/) (Zhang et al., Findings 2026)

• Prism: Persistent Faceted File Memory for Cost-Efficient Coding Agents (Zhang et al., Findings 2026)

ACL

• Ziyun Zhang, Yihao Liu, Yuzhe Zhang, and Huaqian Cai. 2026. Prism: Persistent Faceted File Memory for Cost-Efficient Coding Agents. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 27518–27531, Budapest, Hungary. Association for Computational Linguistics.