Pricing

from $3.00 / 1,000 results

Try for free

Go to Apify Store

Github AI Coding Task Dataset Builder

Try for free

Turn GitHub issues into structured coding-agent tasks with repository context, relevant files, build commands, and ready-to-use prompts.

Pricing

from $3.00 / 1,000 results

Rating

0.0

(0)

Developer

coolinbex

Actor stats

Bookmarked

Total users

Monthly active users

3 days ago

Last modified

Categories

Run mode

mode

Optional

Use the safe built-in demo or scan live GitHub repositories and organizations.

Type:string

Default:demo

Options:

demolive

GitHub repositories

repositories

Optional

Repositories as owner/repo or github.com URLs. You can combine these with organizations and a source dataset.

Type:string[]

Max. items:10000

Default:

[]

GitHub organizations

organizations

Optional

Organization logins or github.com organization URLs. Repositories are expanded automatically.

Type:string[]

Max. items:1000

Default:

[]

Source dataset

sourceDatasetId

Optional

Optional Apify dataset containing GitHub repository strings.

Type:string

Dataset repository field

datasetRepositoryField

Optional

Field containing owner/repo or a GitHub repository URL in each source dataset item.

Type:string

Default:repository

GitHub token

githubToken

Optional

Optional for small public scans, strongly recommended for large batches, and required for private repositories. Read-only repository access is sufficient.

Type:string

Issue state

issueState

Optional

Open issues are best for current agent work. Closed/all issues are useful for benchmark and historical datasets.

Type:string

Default:open

Options:

openclosedall

Updated within days

updatedSinceDays

Optional

Only scan issues updated within this many days. Set 0 to disable the time filter.

Type:integer

Minimum:0

Maximum:3650

Default:365

Require any label

includeLabels

Optional

Optional label allow-list. If set, an issue must have at least one listed label.

Type:string[]

Default:

[]

Exclude labels

excludeLabels

Optional

Issues with any of these labels are skipped before scoring.

Type:string[]

Default:

[
  "duplicate",
  "invalid",
  "wontfix",
  "won't fix",
  "question",
  "support",
  "discussion",
  "not planned"
]

Assignment filter

assignmentFilter

Optional

Keep all issues, only assigned issues, or only unassigned issues.

Type:string

Default:any

Options:

anyunassignedassigned

Minimum task score

minTaskScore

Optional

0-100 heuristic confidence that an issue is an actionable coding task. 55 is a balanced default.

Type:integer

Minimum:0

Maximum:100

Default:55

Maximum repositories

maxRepositories

Optional

Maximum unique repositories expanded and processed in one run.

Type:integer

Minimum:1

Maximum:10000

Default:100

Maximum issues per repository

maxIssuesPerRepository

Optional

Maximum non-pull-request issues scanned per repository, ordered by most recently updated.

Type:integer

Minimum:1

Maximum:10000

Default:500

Maximum output tasks

maxTasks

Optional

Stop emitting new task rows after this many accepted coding tasks.

Type:integer

Minimum:1

Maximum:100000

Default:5000

Include forks

includeForks

Optional

Include repositories marked as forks.

Type:boolean

Default:false

Include archived repositories

includeArchived

Optional

Include archived repositories.

Type:boolean

Default:false

Repository context files

maxContextFiles

Optional

Maximum high-value manifests/config/instruction files fetched per repository for stack and command detection.

Type:integer

Minimum:4

Maximum:60

Default:16

Maximum context file bytes

maxContextFileBytes

Optional

Selected context files larger than this are skipped. Secret-like files are never fetched.

Type:integer

Minimum:10000

Maximum:2000000

Default:250000

Relevant files per task

relevantFilesPerTask

Optional

Maximum likely-relevant repository paths attached to each coding task.

Type:integer

Minimum:1

Maximum:30

Default:8

Maximum issue body characters

maxIssueBodyChars

Optional

Maximum issue-body characters stored in each task and generated prompt.

Type:integer

Minimum:500

Maximum:30000

Default:6000

Generate agent prompt

generateAgentPrompt

Optional

Generate a ready-to-use task prompt for Codex, Cursor, Claude Code, and other coding agents.

Type:boolean

Default:true

Save repository contexts

saveRepositoryContexts

Optional

Save compact per-repository context JSON files in the run key-value store.

Type:boolean

Default:false

Repository concurrency

repositoryConcurrency

Optional

Repositories processed in parallel. Keep conservative to reduce GitHub secondary-rate-limit risk.

Type:integer

Minimum:1

Maximum:10

Default:2

Context-file concurrency

fileConcurrency

Optional

Selected repository blobs fetched in parallel inside each repository.

Type:integer

Minimum:1

Maximum:20

Default:5

Fail run on repository error

failOnRepositoryError

Optional

Useful for CI. Successful task rows and summary are saved first, then the run fails if any repository failed.

Type:boolean

Default:false

GitHub Repo to RAG

bindler/github-repo-to-rag

Turn any public GitHub repository into a clean dataset of code and documentation files. One download, no API token, no rate limits. Built for AI coding assistants and RAG.

Neil Sangwaiya

GitHub Repository Scraper

koreyoshi/github-repo-scraper

Mr-chen

GitHub Repository Search — Repos to JSON

oblanceolate_mandola/github-repo-search

Search GitHub repositories by keyword. Stars, language, last push, open issues, forks as JSON for developer-tooling & research AI agents. $3 per 1,000, no coding.

Hassan Hashish

GitHub Repository Scraper

vulnv/github-repository-scraper

Scrape and extract GitHub repository data, metadata, statistics, stars, forks, issues, and project information from multiple repositories at once.

VulnV

Github Repository Analyzer

actually_good_at_this/apify-github-repository-analyzer

GitHub Repository Analyzer extracts comprehensive repository metrics using the official GitHub API: stars, forks, watchers, contributors, commit activity, and issues/PRs.

john Y

GitHub Repository Details

pepeschuster/github-repo-details

Full stats for specific GitHub repos — stars, forks, issues, license, topics.

Pepe Schuster

GitHub Repository Search API

determined_alghoza_2aj/github-repo-search

Search GitHub for repositories matching a term and get the owner, repository name, description, star count and URL of each result, using only the official GitHub API.

ajay shah

GitHub Issues Scraper

parseforge/github-issues-scraper

Scrapes GitHub Issues from any public repository URL. Returns each issue as a flat row with title, state, labels, assignee, body text, and timestamps. No API key required.

ParseForge

Gitingest: GitHub to LLM Context

gauzy_synthesizer/gitingest-repo-to-llm

Turn any GitHub repository into a single text file optimized for LLMs (ChatGPT, Claude, DeepSeek). Perfect for RAG pipelines, code debugging, and AI context extraction.

DAANISH MANSURI

Prompt Context Builder

wheat_tourist/prompt-context-builder

Build AI-ready context files for any company, product, or topic in seconds. Automatically collects from 10+ sources — website, docs, news, GitHub, Reddit, FAQs, pricing, changelog, careers, and social media — then exports one clean context.md, ready to paste directly into ChatGPT, Claude, or Gemini.