Datasets:

logitly-ai
/

flash-coding-dpo-v1

id stringlengths 11 35	mode stringclasses 3 values	toolset listlengths 1 5	workdir stringlengths 13 38	init_md stringlengths 0 45.7k	chosen listlengths 4 330	rejected listlengths 2 284
vscode-3898637169	work	[ "edit", "glob", "grep", "read", "write" ]	/home/chef/vscode	"# VS Code (Code - OSS) Agent Documentation\n\n## Project Overview\n\nThis is the **Visual Studio Co(...TRUNCATED)	[{"role":"user","content":[{"text":"使用螢幕閱讀器搭配快速開啟對話框時，僅會朗(...TRUNCATED)	[{"role":"user","content":[{"text":"使用螢幕閱讀器搭配快速開啟對話框時，僅會朗(...TRUNCATED)
linux-1613442693	plan	[ "edit", "glob", "grep", "read", "write" ]	/home/supervisor/linux		[{"role":"user","content":[{"text":"Improve the RPC debugging output in rpc_show_tasks() to display (...TRUNCATED)	[{"role":"user","content":[{"text":"Improve the RPC debugging output in rpc_show_tasks() to display (...TRUNCATED)
vault-1518863370	work	[ "bash", "edit", "read", "write" ]	/home/security/vault	"# Agent Documentation for HashiCorp Vault\n\n## Project Overview\n\nHashiCorp Vault is a secrets ma(...TRUNCATED)	[{"role":"user","content":[{"text":"Activer la propagation du contexte d'ID de snapshot via la commu(...TRUNCATED)	[{"role":"user","content":[{"text":"Activer la propagation du contexte d'ID de snapshot via la commu(...TRUNCATED)
linux-3948999277	work	[ "bash", "edit", "read", "write" ]	/home/puppet/linux	"# Linux Kernel Source Repository\n\nThis is the Linux kernel source code (version 4.6-rc2). This do(...TRUNCATED)	[{"role":"user","content":[{"text":"Обновите файл MAINTAINERS, чтобы отразит(...TRUNCATED)	[{"role":"user","content":[{"text":"Обновите файл MAINTAINERS, чтобы отразит(...TRUNCATED)
istio-4094291126	plan	[ "edit", "glob", "grep", "read", "write" ]	/home/supervisor/istio	"# Istio Repository Agent Guide\n\nThis document provides guidance for coding agents operating in th(...TRUNCATED)	[{"role":"user","content":[{"text":"istioctl 을 통해 설치할 때 `variant` 와 같은 전역 (...TRUNCATED)	[{"role":"user","content":[{"text":"istioctl 을 통해 설치할 때 `variant` 와 같은 전역 (...TRUNCATED)
three.js-1533073131	plan	[ "bash" ]	/home/ariel782/three.js	"# Three.js Codebase Guide\n\nThree.js is a JavaScript 3D library that provides <canvas>, <svg>, CSS(...TRUNCATED)	[{"role":"user","content":[{"text":"تحسين وثائق فئات LoadingManager و DefaultLoadingM(...TRUNCATED)	[{"role":"user","content":[{"text":"تحسين وثائق فئات LoadingManager و DefaultLoadingM(...TRUNCATED)
rust-86062193	work	[ "bash", "edit", "read", "write" ]	/home/puppet/rust	"# AGENTS.md - Rust Compiler Repository Guide\n\n## Overview\n\nThis is the **Rust programming langu(...TRUNCATED)	[{"role":"user","content":[{"text":"I need a way to safely run a closure and catch if it panics, ret(...TRUNCATED)	[{"role":"user","content":[{"text":"I need a way to safely run a closure and catch if it panics, ret(...TRUNCATED)
vscode-2273216232	work	[ "bash", "edit", "read", "write" ]	/home/cloud-user/vscode	"# VS Code Agents Instructions\n\nThis file provides instructions for AI coding agents working with (...TRUNCATED)	[{"role":"user","content":[{"text":"The diff editor's resize sash has a visual glitch where the bott(...TRUNCATED)	[{"role":"user","content":[{"text":"The diff editor's resize sash has a visual glitch where the bott(...TRUNCATED)
pytorch-195512101	plan	[ "bash" ]	/home/park/pytorch		[{"role":"user","content":[{"text":"Add support for the Bcube allreduce algorithm to the gloo allred(...TRUNCATED)	[{"role":"user","content":[{"text":"Add support for the Bcube allreduce algorithm to the gloo allred(...TRUNCATED)
rails-496514688	plan	[ "bash" ]	/home/fedora/rails		[{"role":"user","content":[{"text":"Refactor the counter_cache code to use `each` instead of `map` w(...TRUNCATED)	[{"role":"user","content":[{"text":"Refactor the counter_cache code to use `each` instead of `map` w(...TRUNCATED)

End of preview. Expand in Data Studio

Flash Coding

Uploading these datasets to proprietary AI (e.g., asking a question that includes them) can get your account suspended: the AI vendor can misclassify the chain of thought in these records as reverse engineering or duplicating model outputs.

Provider: logitly.ai - Agentic data you can trust

Flash Coding is a dataset of agentic coding traces: an AI coding agent working on tasks in real source repositories.

Every trace is produced under a controlled runtime:

Sandboxed environment, pinned repository, fixed working directory
A restricted toolset and operating mode: work, plan, or cplan

An automated verification protocol validates every record before it enters the dataset.

Read the full specification: specification.pdf.

The dataset has two parts that share a single record schema:

Part	Content	Training method
SFT	One demonstration trace per task	Supervised fine-tuning
DPO	A chosen and a rejected trace of the same task	Direct preference optimization

Trace Format

A trace is a list of messages in the style of the Anthropic Messages API: each message has a role (user or assistant) and a content, a list of typed blocks. Messages strictly alternate between the two roles, starting with user and ending with assistant.

Block	Content
`text`	The task in the first user message; the assistant's narration between tool calls, or the final summary.
`thinking`	The agent's private reasoning for the turn. Plain text, never encrypted or summarized.
`tool_use`	A request to run one tool from `toolset`; `input` is a JSON object.
`tool_result`	The outcome of one `tool_use`, matched by id; `is_error` marks a failed call.

Fixed shapes:

The first user message opens with one text block: the task. Every later user message carries only tool_result blocks, one per tool_use of the preceding assistant message.
Each assistant message holds one thinking block, then any mix of tool_use and text blocks in generation order. The last message closes the trace with a text block: a summary written for the user.

What a trace does not contain:

System prompt. Not recorded: the trace begins with the task itself.
Media. The four block types above are the only kinds; there are no image or document blocks.
Not distillation. No signature and no redacted_thinking: the traces are not distilled from Anthropic models.
Tool schemas. The trace carries the calls, not the definitions: toolset names the tools.

Tools

Tool	Description
`glob`	Match paths against a glob pattern.
`grep`	Search file contents using regex.
`read`	Read contents from a file.
`write`	Write contents to a file, creating or overwriting it entirely.
`edit`	Replace a specific string in a file with new content.
`bash`	Execute a shell command.

Usage

SFT: each record carries a single trace, used for supervised training / fine-tuning.
DPO: each record carries a chosen and a rejected trace of the same task, used for preference training: the model learns to prefer chosen over rejected.

Characteristics

System prompt. Not given; it can be restored from the context fields.
Thinking. Every assistant message carries a thinking block, so the traces carry the complete chain of thought. At training time, a thinking block can be dropped when it is short, so the same traces also serve adaptive thinking.
Language. Tasks are written in multiple languages. In a positive trace, the assistant replies in the task's language; the language of thinking is not defined.
Error flag. is_error=true marks a call that did not exit 0, which can be the ordinary outcome of a call. Keeping these results is deliberate: the model learns to recover from failures over long-horizon work.
Plan mode. In plan mode, writing files is strictly forbidden, and /tmp is not writable either; the one exception is /dev/null. The mode gates use, not existence: the toolset is the same in every mode.

Evaluation

Flash Coding is training data: it carries no scores and no task-success labels, so the effect of training on it cannot be read off the data itself. Measure the effect on external benchmarks; we recommend DeepSWE and Terminal-Bench. The dataset shares no tasks with either benchmark: zero overlap, zero contamination.

Contact

Need a dataset tailored to your tasks? Email research@logitly.ai

Downloads last month: 31