Your projects, on autopilot.

A free, self-hosted project tracker and changelog monitor — snapshots websites, follows GitHub releases and milestones, watches RSS/Atom feeds, flags what matters with priority keyword rules, and summarizes what happened with local AI (Ollama). No cloud, no API bills, your machine.

Everything you watch, one priority-sorted feed

Website snapshots

Capture project pages for offline reading. Scheduled checks detect changes — every change becomes a snapshot version plus a diff in your updates feed. Feeds are discovered automatically when a site publishes one.

GitHub tracking

New releases, README changes and commits per repo — plus milestones and issue-label watching. Major version bumps and stable releases are escalated automatically; no keywords needed.

RSS & Atom feeds

Feeds are first-class sources — often the most reliable way to track a project's changelog. New entries land in the same feed, with the same rules.

Priority keyword rules

Word-boundary matching with negation and priorities: flag anything mentioning kernel, linux as critical. With AI on, a semantic pass catches related text that never uses the exact words.

Local AI (optional)

Change summaries, importance classification, goal extraction and category assignment — via local Ollama (bundled in Docker), OpenAI-compatible, Anthropic or MCP. Fully functional without any AI.

Dashboard & search

Projects grouped by AI-assigned category, activity sparklines, unread counters, digest windows (1d/7d/30d), full-text search over everything, JSON backup/restore.

Screenshots

ProjectPulse dashboard with project cards, activity sparklines and needs-attention strip
Dashboard
Updates feed with priority-sorted updates and AI summaries
Updates feed
Offline website snapshot rendered in a sandboxed iframe with version history
Website snapshot
GitHub tracking: releases, README diffs, milestones per repo
GitHub tracking
Keyword rules with word boundaries, negation and priorities
Keyword rules
AI model manager: Ollama model catalog with one-click download
AI model manager

Run it in 30 seconds

Any Docker host — x86-64 or arm64. AI (local Ollama) comes bundled; the default model downloads automatically on first boot. All your data stays on your machine.

# Docker Compose (bundled Ollama + one-shot model pull)
git clone https://github.com/Pantera87/ProjectPulse
cd ProjectPulse && docker compose up -d
# open http://localhost:4701

# …or just the app (bring your own Ollama / use a remote provider)
docker run -d --name projectpulse -p 4701:4701 -v pp_data:/data \
  pantera87/projectpulse:latest

Then add your first project: a URL, owner/repo or a feed. Set AUTH_PASSWORD to enable a login screen, WEBHOOK_URL to forward updates to ntfy/Pushover, or leave AI off entirely — the app is fully functional without it. Full configuration →

Built in the open

ProjectPulse is MIT-licensed open source, built with Next.js, React and SQLite. Found a bug or want a feature? Open an issue or send a pull request.