SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.00623 · 2026

TraceSIR: A Multi-Agent Framework for Structured Analysis and Reporting of Agentic Execution Traces

Jie Tang, Yidong Wang, Yukuo Cen, Linfeng Zhang, Cunxiang Wang, Lindong Wu, Haoke Zhang, Jiayi Gui, Zhuoer Feng, Wenbo Yu, Shu-Xun Yang, Dayong Yang, and 3 more

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 1 failure of our instrument; 1 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2603.00623.json · badge: /api/ran/2603.00623.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
SHU-XUN/TraceSIR community
Papers with Code archive
3 ran · 1 no recorded run
FunctionStatusWhere it lives
LLMAgentAPI Ran
ran on a synthesized input · constructed an object; no result was computed
SHU-XUN/TraceSIR/pipeline.py
code served (permissive licence) · get_code("535bc1af3796913b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LLMAgentAPI: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
from unittest.mock import Mock, patch, MagicMock

# Set up mock environment variables
os.environ['LLM_MODEL_ID'] = 'gpt-3.5-turbo'
os.environ['LLM_API_KEY'] = 'sk-test-key-12345'
os.environ['LLM_BASE_URL'] = 'https://api.openai.com/v1'
os.environ['LLM_TIMEOUT'] = '600'

# Create mock OpenAI client
mock_client = MagicMock()
mock_response = MagicMock()
mock_response.choices = [MagicMock()]
mock_response.choices[0].message.content = "Test response from LLM"
mock_response.usage = MagicMock(prompt_tokens=10, completion_tokens=5)
mock_client.chat.completions.create.return_value = mock_response

# Patch OpenAI to return our mock client
with patch('openai.OpenAI') as mock_openai_class:
    mock_openai_class.return_value = mock_client

# Define arguments for LLMAgentAPI.__init__
ARGS = {
    'model': 'gpt-3.5-turbo',
    'apiKey': 'sk-test-key-12345',
    'baseUrl': 'https://api.openai.com/v1',
    'timeout': 600
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
openai — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LLMAgentAPI — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LLMAgentAPI output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
LLMAgentToolAPI Ran
ran on a synthesized input · constructed an object; no result was computed
SHU-XUN/TraceSIR/pipeline.py
code served (permissive licence) · get_code("8a1492ab96ed6b30")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LLMAgentToolAPI: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
from unittest.mock import Mock, MagicMock, patch

# Set up environment variables to avoid relying on .env file
os.environ['LLM_MODEL_ID'] = 'gpt-3.5-turbo'
os.environ['LLM_API_KEY'] = 'sk-test-key-12345'
os.environ['LLM_BASE_URL'] = 'https://api.openai.com/v1'
os.environ['LLM_TIMEOUT'] = '600'

# Create mock messages and tools
messages = [
    {'role': 'user', 'content': 'What is the weather today?'},
    {'role': 'assistant', 'content': 'I can help with that.'}
]

tools = [
    {
        'type': 'function',
        'function': {
            'name': 'get_weather',
            'description': 'Get the weather for a location',
            'parameters': {
                'type': 'object',
                'properties': {
                    'location': {'type': 'string', 'description': 'City name'}
                },
                'required': ['location']
            }
        }
    }
]

ARGS = {
    'model': 'gpt-3.5-turbo',
    'apiKey': 'sk-test-key-12345',
    'baseUrl': 'https://api.openai.com/v1',
    'timeout': 600
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
openai — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LLMAgentToolAPI — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LLMAgentToolAPI output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_logger Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
SHU-XUN/TraceSIR/pipeline.py
code served (permissive licence) · get_code("509d3b2efc554c1d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned function. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import queue
from unittest.mock import MagicMock

# Mock the JOB_LOG_QUEUES since it's imported from job_runtime
import sys
from types import ModuleType

job_runtime = ModuleType('job_runtime')
job_runtime.JOB_LOG_QUEUES = {}

sys.modules['job_runtime'] = job_runtime

# Create test data
logs = []
job_id = "test_job_001"

# Create a queue for this job_id
q = queue.Queue()
job_runtime.JOB_LOG_QUEUES[job_id] = q

ARGS = {
    "logs": logs,
    "job_id": job_id,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
job_runtime, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
process_trace_folder Not yet run
unverified (harvested, no recorded run)
SHU-XUN/TraceSIR/pipeline.py
code served (permissive licence) · get_code("8208bbf74f3b0d60")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 4 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Agentic systems augment large language models with external tools and iterative decision making, enabling complex tasks such as deep research, function calling, and coding. However, their long and intricate execution traces make failure diagnosis and root cause analysis extremely challenging. Manual inspection does not scale, while directly applying LLMs to raw traces is hindered by input length limits and unreliable reasoning. Focusing solely on final task outcomes further discards critical behavioral information required for accurate issue localization. To address these issues, we propose TraceSIR, a multi-agent framework for structured analysis and reporting of agentic execution traces. TraceSIR coordinates three specialized agents: (1) Struc-tureAgent, which introduces a novel abstraction format, TraceFormat, to compress execution traces while preserving essential behavioral information; (2) InsightAgent, which performs fine-grained diagnosis including issue localization, root cause analysis, and optimization suggestions; (3) ReportAgent, which aggregates insights across task instances and generates comprehensive analysis reports. To evaluate TraceSIR, we construct TraceBench, covering three real-world agentic scenarios, and introduce ReportEval, an evaluation protocol for assessing the quality and usability of analysis reports aligned with industry needs. Experiments show that TraceSIR consistently produces coherent, informative, and actionable reports, significantly outperforming existing approaches across all evaluation dimensions. Our project and video are publicly available at https://github.com/SHU-XUN/TraceSIR.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2603.00623")
get_code_for_paper("2603.00623")
have("2603.00623")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2603.00623.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2603.00623.svg)](https://syntology.ai/paper/2603.00623)

Connect an agent — have() is free.