owenshuo/open-bounty-radar: open Bounty Radar is a small CLI for disco

Open Bounty Radar is a small CLI for discovering and monitoring paid open-source issues, bounty-style GitHub issues, and competitive pull request opportunities.

It helps turn scattered bounty research into a repeatable workflow:

configure repos -> scan bounty issues -> score candidates -> inspect competition -> watch submitted PRs

It is designed for developers who want to find issues that are:

still open
clearly paid or bounty-backed
not already crowded by many pull requests
practical to review quickly from a Linux-friendly development workflow

Who this is for

Independent developers who want a practical bounty radar instead of manually refreshing GitHub searches.
Open-source contributors who need to avoid already-crowded or already-solved issues.
Small teams tracking submitted bounty PRs, maintainer comments, checks, and competitor movement.
Builders who want a local-first workflow with explainable scoring and no hosted service requirement.

Why this exists

Paid OSS issues are scattered across GitHub labels, issue bodies, platform comments, and project-specific conventions such as /bounty $6000. Developers waste time opening issues that are already solved, crowded, closed, or missing payment details.

This project turns that manual research into a repeatable scan and report.

Project Status

Open Bounty Radar is an early-stage OSS project with a working 0.1.x CLI surface, documented configuration, offline demo fixtures, automated tests, CI, and a release checklist. The current goal is to make bounty-related OSS contribution more transparent and less duplicative, not to encourage rushed or spammy pull requests.

Suggested repository topics: open-source, bounty, github, cli, developer-tools, triage, pull-requests, oss.

Current MVP

Scans configured GitHub repositories and GitHub-wide issue searches with GitHub Search API
Detects bounty amounts such as $250, $6k, /bounty $6000, and 250 USDC
Filters closed issues by default
Searches for pull requests that mention each issue
Reads GitHub issue timeline cross-references to catch linked PRs search can miss
Enriches competing PRs with draft/merged state, maintainer review signals, checks, and strength labels
Scores candidates by bounty amount, freshness, open state, and PR competition
Adds recommendation and risk tags such as strong, risky, crowded, and special-requirements
Assigns risk severity levels so high-risk blockers stand out from minor warnings
Adds action labels such as act-now, watch, manual-review, and skip
Highlights top candidates so the report opens with the most actionable issues
Shows linked PR competition details with state, update date, and detection source
Writes Markdown, JSON, static HTML, and dashboard reports
Supports GitHub search presets such as bounty, external, recent, low-competition, crypto-bounty, reward, and amounts
Provides an interactive static dashboard with filters, search, copyable issue URLs, and candidate detail pages
Includes Algora and Opire adapter foundations with optional GitHub issue enrichment
Can extract GitHub-linked bounty listings from simple live HTML sources
Adds local AI-style assessments with verdict, confidence, next steps, likely files, and abandon conditions
Adds PR readiness checks for reproduction, scope, requirements, bounty, and competition risk
Exports watchlist suggestions so promising candidates can be turned into submitted-PR monitoring entries
Maintains optional local workspace state for candidate status and notes
Adds watch next-action hints such as reply, fix-ci, revise, and claim-or-confirm
Includes a GitHub Actions workflow template for scheduled runs
Watches submitted pull requests for merge/close state, checks, reviews, and maintainer comments
Stores state snapshots and detects meaningful changes between runs
Sends Telegram or generic webhook notifications for detected changes

Quick Start

git clone https://github.com/owenshuo/open-bounty-radar.git
cd open-bounty-radar
npm test
npm run init
npm run doctor
npm run validate
npm run radar

npm run init creates local config files, then npm run radar reads bounty-radar.json and runs both configured jobs:

scan open bounty candidates
watch already-submitted pull requests

Markdown, JSON, static HTML, and dashboard reports are written to reports/ by default.

npm run doctor checks your Node.js version, config files, output directories, token setup, and GitHub API connectivity.

npm run validate checks the radar config and referenced scan/watch configs without making GitHub API calls.

npm run init creates local config files from the examples:

bounty-radar.json
bounty-radar.config.json
bounty-radar.watchlist.json

Local config files are ignored by git so you can customize them safely.

5-Minute Tour

Run npm run init to create local config files.
Edit bounty-radar.config.json and add repositories or GitHub-wide search queries you care about.
Run npm run doctor to confirm the local environment and GitHub API access.
Run npm run validate to catch config mistakes without using GitHub API calls.
Run npm run radar to create Markdown, JSON, and HTML reports.
Open reports/bounty-report.html and start with the Top Candidates section.
Add submitted PRs to bounty-radar.watchlist.json, then keep using npm run radar to monitor them.

See Demo Output for a compact example of the generated reports.

Want to try the UI without a GitHub token or live API calls?

Open the hosted demo dashboard or run it locally:

npm run demo:offline
npm run serve

Then open reports/demo-dashboard.html. The fixture data lives in examples/fixtures/demo-listings.json, so this path is safe for hosted demos, screenshots, and release smoke tests.

Open Bounty Radar demo dashboard

Guides

Outputs

Open Bounty Radar can write:

Markdown reports
JSON reports
static HTML reports
scan dashboard
watch dashboard
CSV exports
JSONL exports
action plans
watchlist suggestion JSON
workspace state JSON
history JSONL

To run each job separately:

npm run scan
npm run watch

To run a broader discovery pass that keeps more candidates for review while still applying risk labels:

export GITHUB_TOKEN=github_pat_xxx
npm run scan:discovery

Discovery mode uses more GitHub Search queries than the compact example, so a token is strongly recommended.

To validate an example or custom radar config:

node ./bin/open-bounty-radar.js validate --config ./examples/radar.full.json

To create local config files again and overwrite existing ones:

node ./bin/open-bounty-radar.js init --force

To enable Telegram notifications without adding CLI flags, set notifications.telegram.enabled to true in bounty-radar.config.json and/or bounty-radar.watchlist.json, then run:

npm run radar

For higher rate limits, set a GitHub token:

export GITHUB_TOKEN=github_pat_xxx
export TELEGRAM_BOT_TOKEN=123456:your_bot_token
export TELEGRAM_CHAT_ID=123456789
npm run scan

On Windows PowerShell:

$env:GITHUB_TOKEN="github_pat_xxx"
$env:TELEGRAM_BOT_TOKEN="123456:your_bot_token"
$env:TELEGRAM_CHAT_ID="123456789"
npm run scan

Configuration

The recommended one-command entrypoint is the local bounty-radar.json generated by npm run init:

{
  "scan": {
    "enabled": true,
    "config": "./bounty-radar.config.json",
    "out": "./reports/bounty-report.md",
    "json": "./reports/bounty-report.json",
    "html": "./reports/bounty-report.html",
    "dashboard": "./reports/dashboard.html",
    "detailsDir": "./reports/details",
    "watchlistSuggestions": "./reports/watchlist-suggestions.json"
  },
  "watch": {
    "enabled": true,
    "config": "./bounty-radar.watchlist.json",
    "out": "./reports/pr-watch.md",
    "json": "./reports/pr-watch.json",
    "html": "./reports/pr-watch.html"
  }
}

Run it with:

npm run radar

You can disable either section by setting "enabled": false.

Additional example profiles:

examples/radar.minimal.json: scan-only starter profile.
examples/radar.full.json: explicit scan + watch profile for copying.

Development/demo commands that run directly from examples/ are also available:

npm run validate:example
npm run doctor:example
npm run radar:example
npm run dashboard:example
npm run demo:offline
npm run audit
npm run release:check
npm run serve

Inspect one issue deeply:

node ./bin/open-bounty-radar.js inspect --issue-url https://github.com/owner/repo/issues/123 --out ./reports/issue-inspection.md --json ./reports/issue-inspection.json --html ./reports/issue-inspection.html

Batch inspect issues from a text file:

node ./bin/open-bounty-radar.js inspect --issue-list ./issues.txt --out ./reports/issue-batch.md --json ./reports/issue-batch.json

Add --html ./reports/issue-batch.html to create a browser-friendly batch inspection dashboard. Batch HTML also writes linked per-issue detail pages under reports/issue-details/ by default:

node ./bin/open-bounty-radar.js inspect --issue-list ./issues.txt --out ./reports/issue-batch.md --json ./reports/issue-batch.json --html ./reports/issue-batch.html --inspect-details-dir ./reports/issue-details

Scan Config

Create a JSON config:

{
  "githubTokenEnv": "GITHUB_TOKEN",
  "statePath": "./reports/radar-state.json",
  "notifications": {
    "notifyOnFirstRun": false,
    "telegram": {
      "enabled": false,
      "botTokenEnv": "TELEGRAM_BOT_TOKEN",
      "chatIdEnv": "TELEGRAM_CHAT_ID"
    }
  },
  "defaults": {
    "maxIssuesPerQuery": 20,
    "globalMaxIssuesPerQuery": 5,
    "includeClosed": false,
    "linkedPullRequestDetection": "both",
    "competitionDetails": true,
    "competitionDetailLimit": 5
  },
  "workspacePath": "./reports/workspace.json",
  "filters": {
    "minAmount": 100,
    "excludeKeywords": ["marketing", "hardware", "ios only"]
  },
  "repositories": [
    {
      "owner": "Expensify",
      "repo": "App",
      "queries": ["$ in:title,body label:External"]
    }
  ],
  "githubSearches": [
    {
      "name": "global-bounty-labels",
      "queries": ["label:bounty $ in:title,body archived:false"]
    }
  ]
}

Then run:

node ./bin/open-bounty-radar.js scan --config ./examples/config.json --out ./reports/bounty-report.md --json ./reports/bounty-report.json --html ./reports/bounty-report.html

Add --state to compare this run with the previous run:

node ./bin/open-bounty-radar.js scan --config ./examples/config.json --out ./reports/bounty-report.md --state ./reports/radar-state.json

Watching Pull Requests

Watch Config

Create a watchlist:

{
  "githubTokenEnv": "GITHUB_TOKEN",
  "statePath": "./reports/radar-state.json",
  "notifications": {
    "notifyOnFirstRun": false,
    "telegram": {
      "enabled": false,
      "botTokenEnv": "TELEGRAM_BOT_TOKEN",
      "chatIdEnv": "TELEGRAM_CHAT_ID"
    }
  },
  "defaults": {
    "activityLimit": 5
  },
  "pullRequests": [
    {
      "owner": "spaceandtimefdn",
      "repo": "sxt-proof-of-sql",
      "number": 1986,
      "label": "Example watched bounty PR"
    }
  ]
}

Then run:

node ./bin/open-bounty-radar.js watch --config ./examples/watchlist.json --out ./reports/pr-watch.md --json ./reports/pr-watch.json --html ./reports/pr-watch.html

The watch report highlights PRs that need attention because they were closed, have failing checks, or received maintainer/owner activity.

Notifications

Notifications are change-based. The first run creates a baseline state file, then later runs notify only when something meaningful changes.

Telegram environment variables:

export TELEGRAM_BOT_TOKEN=123456:your_bot_token
export TELEGRAM_CHAT_ID=123456789

Run with notifications by turning on Telegram in the JSON config:

{
  "notifications": {
    "telegram": {
      "enabled": true
    }
  }
}

Then use the normal one-command entrypoint:

npm run radar

You can also force notification mode from the CLI:

node ./bin/open-bounty-radar.js watch --config ./examples/watchlist.json --out ./reports/pr-watch.md --state ./reports/radar-state.json --notify

To enable notifications from config instead of passing --notify, set notifications.telegram.enabled to true.

Generic JSON webhooks are also supported:

{
  "notifications": {
    "webhook": {
      "enabled": true,
      "urlEnv": "OPEN_BOUNTY_RADAR_WEBHOOK_URL"
    }
  }
}

The webhook payload includes the run kind, generation time, compact digest text, and the structured change list.

Discord and Slack incoming webhooks can use the same digest:

{
  "notifications": {
    "discord": {"enabled": true},
    "slack": {"enabled": true}
  }
}

Set DISCORD_WEBHOOK_URL or SLACK_WEBHOOK_URL in the environment.

Notification rules let routine scans stay quieter:

{
  "notifications": {
    "rules": {
      "minSeverity": "medium",
      "actions": ["act-now", "manual-review"],
      "minAmount": 250,
      "competitionRisks": ["none", "low"]
    }
  }
}

You can also use presets: quiet, aggressive, high-value-only, or low-competition-only.

Workspace state can be merged back from a dashboard export:

node ./bin/open-bounty-radar.js scan --workspace ./reports/workspace.json --workspace-import ./exports/workbench.json

You can also summarize or merge workspace state directly:

node ./bin/open-bounty-radar.js workspace --workspace ./reports/workspace.json --workspace-import ./exports/workbench.json --out ./reports/workspace-summary.md --json ./reports/workspace.json

The dashboard also supports importing a previously exported workspace JSON file directly in the browser. Use Export Workspace to copy the current local workbench state, save it as JSON if desired, then Import Workspace to restore status and notes on another generated dashboard.

Config Schema

A formal JSON Schema is included at schema/open-bounty-radar.schema.json. It covers the top-level radar config plus scan and watch config shapes, including GitHub repositories, GitHub-wide searches, Algora/Opire listing sources, notifications, workspace paths, and report outputs.

Release Check

Before tagging or sharing a release candidate, run:

npm run release:check

It runs tests, example validation, the offline demo scan, package audit, and git diff --check.

Scoring

The score is intentionally simple:

higher bounty amount improves score
recently updated issues score higher
open issues score higher than closed ones
each linked or mentioned PR reduces score
strong competing PRs with approvals, passing checks, or merged state increase risk
AI-style assessments summarize whether to start, watch, manually review, or abandon

Linked PR detection supports search, timeline, or both. The default both mode merges GitHub Search results with issue timeline cross-references, then de-duplicates by PR URL.

This is not meant to decide for you. It is meant to triage quickly.

Roadmap

The short version:

better setup diagnostics
richer HTML reports
Algora and Opire adapters
GitHub Actions scheduled reports
maintainer assignment and winner-detection heuristics
local web dashboard

See the full Roadmap.

Ethics

This project should help contributors find suitable work and reduce duplicate effort. It should not be used to spam maintainers, mass-generate low-quality pull requests, or bypass project contribution rules.

Always read the issue, reproduce the bug, follow the project process, and respect maintainers.

owenshuo/open-bounty-radar: open Bounty Radar is a small CLI for discovering and monitoring paid open-source

Who this is for

Why this exists

Project Status

Current MVP

Quick Start

5-Minute Tour

Guides

Outputs

Configuration

Scan Config

Watching Pull Requests

Watch Config

Notifications

Config Schema

Release Check

Scoring

Roadmap

Ethics

Comments

Who this is for

Why this exists

Project Status

Current MVP

Quick Start

5-Minute Tour

Guides

Outputs

Configuration

Scan Config

Watching Pull Requests

Watch Config

Notifications

Config Schema

Release Check

Scoring

Roadmap

Ethics

Comments

Related Posts

Wumpus-multitool: wumpus Multitool — Le meilleur outil tout-en-un Discord (Self-bot & Token Tools) avec

edgenuitybot/exodusbot: see the full, regularly updated list at exodusbot.net/features. All 30+ features

strmax195-hue/docktui: lightweight self-hosted alternative for tinkerers

mansoor-mamnoon/codeglance: no install. No config. No API keys.