Shuo Li

—

Unclaimed · AI-structured profileOpenAlex

Early-career software engineering researcher: LLM-agent benchmarks, repo governance, trust-native agent systems

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-17) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

Insufficient data: the snapshot covers only 3 works recorded in 2026 and no verified full publication history, so overall output volume cannot be scored. This axis would measure cumulative and per-year publication counts across venues and years.

Impact—

Impact is not scored: citation coverage for this OpenAlex record is essentially empty, with h-index and i10-index both recorded as 0. This axis would measure citation-based influence such as citations per paper, h-index, and uptake over time.

Focus100/100

Focus scores a perfect 100: all five OpenAlex topics on this record sit in one cluster — Software Engineering Research (2) and Software Engineering Techniques and Practices (2), plus Software Testing and Debugging, Access Control and Trust, and Security and Verification (1 each) — and the single field-tagged paper falls in software engineering (LLM agent evaluation). The two 2026 catalog papers extend the same agentic-engineering line.

Collaboration56/100

Collaboration scores 56, drawn from 16 distinct coauthors in the OpenAlex record. Twelve recur as frequent collaborators — including Jiajun Wu, Leixin Sun, Zihan Tan, and Yang Liu — suggesting a stable, sizeable team pattern; both 2026 catalog papers are multi-author works with Shuo Li in a middle authorship position rather than leading.

Activity40/100

Activity scores 40: the year-by-year counts list only 2026, with 3 works, and the record shows 3 works over the last three years — a young, front-loaded profile. There is no visible multi-year publication cadence yet, so this score reflects a short observable window rather than a decline.

Reproducibility55/100

Reproducibility scores 55 — an average of 55 over just 1 Evolve-assessed paper. SWE-Prometheus is rated medium readiness: the benchmark ships rigorous controls (no-op, template, and gate ablation), but its single-rollout, LLM-judged scoring limits measurement precision, so the evidence is real yet thin.

Research summary

Shuo Li is an early-stage software engineering researcher whose public record, resolved from OpenAlex bibliographic data, is concentrated almost entirely on agentic and LLM-driven software engineering. All five OpenAlex topics on the record — Software Engineering Research, Software Engineering Techniques and Practices, Software Testing and Debugging, Access Control and Trust, and Security and Verification — sit in one tightly related cluster, earning a maximal focus score of 100. The visible output is recent and compact: 3 works recorded in 2026, including two arXiv preprints in the Evolve catalog. SWE-Prometheus (2026) proposes a benchmark for measuring how LLM coding agents improve engineering governance in real-world repositories, with no-op, template, and gate-ablation controls; Evolve rates it medium reproducibility readiness (score 55 from 1 assessed paper) while noting that single-rollout LLM-judged scoring caps precision. AgentKernel (2026) proposes a trust-native agentic operating system, extending the same trust and governance agenda toward systems infrastructure. Collaboration is substantial for an early record — 16 distinct coauthors, with 12 recurring collaborators — though Shuo Li appears in middle authorship positions on both catalog papers. Citation-based impact cannot yet be assessed: h-index and i10-index are recorded as 0, and the output and impact dimensions are marked missing. With no verified institution, country, or ORCID, this remains an early-stage profile.

Collaboration pattern

Shuo Li's record lists 16 distinct coauthors, and 12 recur as frequent collaborators — including Jiajun Wu, Leixin Sun, Zihan Tan, Yitao Liu, Jiaru Qian, Shanghaoran Quan, Chuangxin Zhao, Yangxu Liao, Yang Liu, Bin Chong, Guancheng Wan, and Zhenhua Zou — pointing to a stable, sizeable team. Both 2026 catalog papers are multi-author efforts with Shuo Li in a middle authorship position, consistent with contributing within a larger group rather than solo or senior-led work.

Research keywords

LLM coding agentsAgent benchmarkingSoftware engineering governanceRepository-level evaluationBehavior gatesMutation testingAgentic operating systemsTrust and access controlSecurity and verificationSoftware testing and debugging

Highlights

SWE-Prometheus: benchmarking governance gains from LLM coding agents
A 2026 arXiv benchmark measuring engineering governance improvements in real-world repositories, with rigorous controls — no-op, template, and gate ablation — though single-rollout, LLM-judged scoring caps measurement precision; Evolve rates reproducibility readiness medium (score 55 from 1 assessed paper).
AgentKernel: proposing a trust-native agentic operating system
A second 2026 arXiv preprint framing agent infrastructure as a trust-native operating system, connecting the record's Access Control and Trust and Security and Verification topics; Evolve has not yet published a verdict, keywords, or reproducibility rating for it.
A maximally focused early record
All five OpenAlex topics — Software Engineering Research (2), Software Engineering Techniques and Practices (2), Software Testing and Debugging, Access Control and Trust, and Security and Verification (1 each) — form one coherent cluster, yielding a focus score of 100 on an otherwise sparse record.

Timeline

2026First visible publication year: OpenAlex records 3 works, all falling within 2026, with no earlier output on the record.
2026Co-authors SWE-Prometheus (arXiv), a benchmark measuring how LLM coding agents improve engineering governance in real-world repositories; it carries the record's only field tag, software engineering (LLM agent evaluation).
2026Co-authors AgentKernel (arXiv), proposing a trust-native agentic operating system, aligning with the record's Access Control and Trust and Security and Verification topic clusters.
2026Evolve publishes its verdict on SWE-Prometheus: rigorous controls (no-op, template, gate ablation), but single-rollout LLM-judged scoring caps precision, with medium reproducibility readiness.

Papers in the Evolve catalog

2026-09-24
SWE-Prometheus: Measuring Engineering Governance Improvements in Real-World Repositories

Frequent co-authors

Jiajun Wu×1 Leixin Sun×1 Zihan Tan×1 Yitao Liu×1 Jiaru Qian×1 Shanghaoran Quan×1 Chuangxin Zhao×1 Yangxu Liao×1 Yang Liu×1 Bin Chong×1 Guancheng Wan×1 Zhenhua Zou×1

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometrics and topic labels come from OpenAlex only; no ORCID, institution, or country is verified, and "Shuo Li" is a common name, so this profile is name-resolved from public bibliographic data.
Citation indicators are absent (h-index and i10-index recorded as 0), so the output and impact dimensions are unscored rather than genuinely low.
Only one of the two catalog papers carries a full Evolve review; the reproducibility score of 55 rests on a single assessed paper, and AgentKernel has no verdict, keywords, or readiness rating yet.
All observed works date to 2026; there is no multi-year publication history from which to establish trend or cadence.