Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

APISENSOR: Robust Discovery of Web API from Runtime Traffic Logs

Yanjing Yang, Chenxing Zhong, Ke Han, Zeru Cheng, Jinwei Xu, Xin Zhou, He Zhang, Bohan Liu

Abstract

Large Language Model (LLM)-based agents increasingly rely on APIs to operate complex web applications, but rapid evolution often leads to incomplete or inconsistent API documentation. Existing work falls into two categories: (1) static, white-box approaches based on source code or formal specifications, and (2) dynamic, black-box approaches that infer APIs from runtime traffic. Static approaches rely on internal artifacts, which are typically unavailable for closed-source systems, and often over-approximate API usage, resulting in high false-positive rates. Although dynamic black-box API discovery applies broadly, its robustness degrades in complex environments where shared collection points aggregate traffic from multiple applications. To improve robustness under mixed runtime traffic, we propose APISENSOR, a black-box API discovery framework that reconstructs application APIs unsupervised. APISENSOR performs structured analysis over complex traffic, combining traffic denoising and normalization with a graph-based two-stage clustering process to recover accurate APIs. We evaluated APISENSOR across six web applications using over 10,000 runtime requests with simulated mixed-traffic noise. Results demonstrate that APISENSOR significantly improves discovery accuracy, achieving an average Group Accuracy Precision of 95.92% and an F1-score of 94.91%, outperforming state-of-the-art methods. Across different applications and noise settings, APISENSOR achieves the lowest performance variance and at most an 8.11-point FGA drop, demonstrating the best robustness among 10 baselines. Ablation studies confirm that each component is essential. Furthermore, APISENSOR revealed API documentation inconsistencies in a real application, later confirmed by community developers.

APISENSOR: Robust Discovery of Web API from Runtime Traffic Logs

Abstract

Large Language Model (LLM)-based agents increasingly rely on APIs to operate complex web applications, but rapid evolution often leads to incomplete or inconsistent API documentation. Existing work falls into two categories: (1) static, white-box approaches based on source code or formal specifications, and (2) dynamic, black-box approaches that infer APIs from runtime traffic. Static approaches rely on internal artifacts, which are typically unavailable for closed-source systems, and often over-approximate API usage, resulting in high false-positive rates. Although dynamic black-box API discovery applies broadly, its robustness degrades in complex environments where shared collection points aggregate traffic from multiple applications. To improve robustness under mixed runtime traffic, we propose APISENSOR, a black-box API discovery framework that reconstructs application APIs unsupervised. APISENSOR performs structured analysis over complex traffic, combining traffic denoising and normalization with a graph-based two-stage clustering process to recover accurate APIs. We evaluated APISENSOR across six web applications using over 10,000 runtime requests with simulated mixed-traffic noise. Results demonstrate that APISENSOR significantly improves discovery accuracy, achieving an average Group Accuracy Precision of 95.92% and an F1-score of 94.91%, outperforming state-of-the-art methods. Across different applications and noise settings, APISENSOR achieves the lowest performance variance and at most an 8.11-point FGA drop, demonstrating the best robustness among 10 baselines. Ablation studies confirm that each component is essential. Furthermore, APISENSOR revealed API documentation inconsistencies in a real application, later confirmed by community developers.
Paper Structure (25 sections, 5 equations, 6 figures, 6 tables, 1 algorithm)

This paper contains 25 sections, 5 equations, 6 figures, 6 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Approach
  4. Overview
  5. Statistical Traffic Denoising and Path Normalization
  6. Two-Stage Clustering with Structural Templates and Semantic Graphs.
  7. Experiment Design
  8. Research Questions
  9. Studied Applications
  10. Noisy Simulation
  11. Selected Baseline
  12. Evaluation Metrics
  13. Result Analysis
  14. Effectiveness (RQ1)
  15. Robustness (RQ2)
  16. ...and 10 more sections

Figures (6)

  • Figure 1: Existing tools (e.g., Optic) cannot reliably distinguish these heterogeneous requests, potentially leading to incomplete and inaccurate API specifications.
  • Figure 2: The workflow of APISensor.
  • Figure 3: Structural template mining via Drain3. Concrete API paths are abstracted into interface-level templates by replacing dynamic identifier segments.
  • Figure 4: Performance Stability of API Endpoint Discovery Methods Across Diverse Projects (PGA, RGA, FGA in %)
  • Figure 5: Robustness of API discovery under increasing noise ratios (Interfere and Lexify).
  • ...and 1 more figures