graphectory-project.github.io/index.html at main · Intelligent-CAT-Lab/graphectory-project.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="description"
        content="Process-Centric Analysis of Agentic Software Systems introduces Graphectory and Langutory for systematic trajectory analysis, online monitoring, and intervention in software engineering agents.">
  <meta name="keywords" content="Graphectory, Langutory, Agentic Software Systems, SWE-agent, OpenHands, Software Engineering Agents, Process-Centric Analysis">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Process-Centric Analysis of Agentic Software Systems</title>

  <!-- Replace with your own icon if available -->
  <!-- <link rel="icon" href="./static/images/favicon.png"> -->

  <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">

  <link rel="stylesheet" href="./static/css/bulma.min.css">
  <link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
  <link rel="stylesheet" href="./static/css/bulma-slider.min.css">
  <link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
  <link rel="stylesheet" href="./static/css/index.css">

  <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
  <script defer src="./static/js/fontawesome.all.min.js"></script>
  <script src="./static/js/bulma-carousel.min.js"></script>
  <script src="./static/js/bulma-slider.min.js"></script>
</head>

<body>

<section class="hero">
  <div class="hero-body">
    <div class="container is-max-desktop">
      <div class="columns is-centered">
        <div class="column has-text-centered">

          <h1 class="title is-1 publication-title">
            Process-Centric Analysis of Agentic Software Systems
          </h1>

          <div class="is-size-5 publication-authors">
            <span class="author-block">Shuyang Liu<sup>1</sup>,</span>
            <span class="author-block">Yang Chen<sup>1</sup>,</span>
            <span class="author-block">Rahul Krishna<sup>2</sup>,</span>
            <span class="author-block">Saurabh Sinha<sup>2</sup>,</span>
            <span class="author-block">Jatin Ganhotra<sup>2</sup>,</span>
            <span class="author-block">Reyhan Jabbarvand<sup>1</sup></span>
          </div>

          <div class="is-size-5 publication-authors" style="margin-top: 0.75rem;">
            <span class="author-block" style="margin-right: 1.5rem;">
              <img src="./static/images/uiuc.svg" alt="UIUC" style="height: 1.2em; vertical-align: middle; margin-right: 0.35rem;">
              <sup>1</sup>University of Illinois Urbana-Champaign
            </span>
            <span class="author-block">
              <img src="./static/images/ibm.svg" alt="IBM" style="height: 1.2em; vertical-align: middle; margin-right: 0.35rem;">
              <sup>2</sup>IBM Research
            </span>
          </div>

          <div class="is-size-5 publication-authors" style="margin-top: 0.6rem;">
            <span class="author-block" style="color: rgb(192, 54, 54);">
              <b>OOPSLA 2026</b>
            </span>
          </div>

          <div class="column has-text-centered" style="margin-top: 1rem;">
            <div class="publication-links">
              <span class="link-block">
                <a href="https://arxiv.org/abs/2512.02393"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <i class="ai ai-arxiv"></i>
                  </span>
                  <span>Paper</span>
                </a>
              </span>

              <span class="link-block">
                <a href="https://github.com/Intelligent-CAT-Lab/Graphectory.git"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <i class="fab fa-github"></i>
                  </span>
                  <span>Code</span>
                </a>
              </span>

              <!-- Optional: update once you have a direct PDF link -->

              <span class="link-block">
                <a href="./static/pdfs/Graphectory_arxiv.pdf"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <i class="fas fa-file-pdf"></i>
                  </span>
                  <span>PDF</span>
                </a>
              </span>

            </div>
          </div>

        </div>
      </div>
    </div>
  </div>
</section>


<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3">Abstract</h2>
        <div class="content has-text-justified" style="font-size: 1.12rem;">
          <p>
            Agentic systems are modern software systems: they consist of orchestrated modules, expose interfaces, and are deployed in software pipelines. Unlike conventional programs, their execution trajectories are inherently stochastic and adaptive to the problems they solve. Evaluation of such systems is often outcome-centric, focusing only on whether the final task succeeds or fails. This narrow view overlooks how agents reason, plan, act, and revise their strategies during execution.
          </p>
          <p>
            We introduce <b>Graphectory</b>, a graph-based representation that systematically encodes the temporal and semantic relations in agent trajectories, enabling process-centric metrics and analyses of workflow quality. Using Graphectory, we analyze 4,000 trajectories from two leading software engineering agents, <b>SWE-agent</b> and <b>OpenHands</b>, paired with four backbone large language models on SWE-bench Verified.
          </p>
          <p>
            Our fully automated analyses reveal that stronger models and richer prompts lead to more complex yet more coherent workflows, while unsuccessful trajectories tend to exhibit chaotic, repetitive, or backtracking behaviors. We further develop a real-time monitoring framework that constructs and analyzes <b>Graphectory</b> and <b>Langutory</b> online to detect process-level issues, notify the agent with diagnostic feedback, and roll back execution when appropriate. Experiments show that online monitoring with intervention improves resolution rates by 6.9%–23.5% on problematic instances while significantly shortening trajectories with near-zero overhead.
          </p>
        </div>
      </div>
    </div>
  </div>
</section>


<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered">
      <div class="column is-four-fifths">

        <h2 class="title is-3 has-text-centered">Interactive Demo</h2>

        <div class="content has-text-justified" style="font-size: 1.12rem; margin-bottom: 1.5rem;">
          <p>
            <b>Graphectory</b> provides interactive visualization and process-centric analysis of agent trajectories, enabling users to inspect how software engineering agents reason, explore, patch, and validate over time.
          </p>
          <p>
            Our current implementation supports trajectories from <b>SWE-agent</b>, <b>OpenHands</b>, and <b>mini-swe-agent</b>. The framework is designed to be extensible, allowing users to adapt it with minimal effort to other agentic systems.
          </p>
        </div>

        <div class="has-text-centered">
          <img src="./static/images/html_demo.png"
               alt="Interactive visualization of Graphectory"
               style="width: 100%; border-radius: 12px;" />
        </div>

        <div class="content has-text-centered" style="margin-top: 1rem; font-size: 1.02rem;">
          <p>
            Interactive visualization of Graphectory for trajectory inspection and process-centric analysis.
          </p>
        </div>

      </div>
    </div>
  </div>
</section>

<section class="section" id="BibTeX">
  <div class="container is-max-desktop content">
    <h2 class="title is-3 has-text-centered">BibTeX</h2>
    <pre><code>@article{liu2025process,
  title={Process-Centric Analysis of Agentic Software Systems},
  author={Liu, Shuyang and Chen, Yang and Krishna, Rahul and Sinha, Saurabh and Ganhotra, Jatin and Jabbarvand, Reyhan},
  journal={arXiv preprint arXiv:2512.02393},
  year={2025}
}</code></pre>
  </div>
</section>


<footer class="footer">
  <div class="columns is-centered">
    <div class="column is-8">
      <div class="content">
        <p style="font-size: 1.05rem;">
          This website is adapted from <a href="https://nerfies.github.io/">Nerfies</a>, licensed under a
          <a rel="license" href="https://creativecommons.org/licenses/by-sa/4.0/">
            Creative Commons Attribution-ShareAlike 4.0 International License
          </a>.
        </p>
      </div>
    </div>
  </div>
</footer>

</body>
</html>