Harken
What it is
Harken tracks keywords, brands, or products across online sources and organizes mentions, sentiment, and themes in a local dashboard. Its sources include Hacker News, Reddit, Mastodon, Bluesky, Stack Overflow, RSS, X, and YouTube. A built-in local lexicon analyzer scores sentiment; optional LLM analysis can label themes using Anthropic, OpenAI, or locally run Ollama. Users can work from a local web dashboard or a terminal CLI. Harken stores its database in SQLite and says it sends no Harken telemetry. Alerts for new negative mentions and configured volume or sentiment thresholds can go to Slack, generic webhooks, or SMTP email. Tracking sends search terms to the selected source APIs, and using Anthropic or OpenAI for theme labels sends a small sample of mention text to that provider. Hacker News and Bluesky work without credentials; other sources may require credentials, feeds, or tokens. The free MIT-licensed software is self-hosted, has no per-seat pricing, and includes API access. Historical data is limited. The quickstart requires Python 3.10 or later, and Docker Compose instructions are included.
For and against
For
- SQLite database is stored locally.
- Sends no Harken telemetry, according to the project.
- Alerts support Slack, webhooks, and SMTP email.
- Optional theme analysis supports locally run Ollama.
Against
- Historical data is limited.
- Several sources require credentials, feeds, or tokens.
- Tracking sends search terms to selected source APIs.
- Requires Python 3.10 or later for quickstart.
The verdict
Harken combines multi-source mention tracking with local storage, sentiment scoring, and configurable alerts. Consider the source setup requirements and data sent to optional external LLM providers before enabling those integrations.
Who it suits
Harken suits indie founders, open-source maintainers, and privacy-conscious teams who want to monitor online mentions in a self-hosted setup. It may also fit users who prefer local storage and a terminal interface alongside a dashboard.
Plans and prices
Free
Open-source MIT · self-hosted · no per-seat pricing
Read 7 Oct 2026 · github.com
On the record
- Alerts
- It supports Slack or generic webhooks and SMTP email alerts for new negative mentions and configured volume or sentiment thresholds.github.com · read 7 Oct 2026
- Analysis
- It scores sentiment with a built-in local lexicon analyzer and groups mentions into themes, with optional LLM analysis.github.com · read 7 Oct 2026
- Authentication
- Optional account mode supports viewer, operator, and admin roles, with salted PBKDF2-HMAC-SHA256 password hashes and expiring opaque sessions.github.com · read 7 Oct 2026
- Compatibility
- The quickstart requires Python 3.10 or later and the repository includes Docker Compose instructions.github.com · read 7 Oct 2026
- Intended users
- The README describes Harken as suited to indie founders, open-source maintainers, and privacy-conscious teams.github.com · read 7 Oct 2026
- Interfaces
- Harken provides a local web dashboard and a terminal CLI.github.com · read 7 Oct 2026
- Optional LLM providers
- Optional theme labeling supports Anthropic, OpenAI, or locally run Ollama.github.com · read 7 Oct 2026
- Privacy detail
- Tracking sends search terms to selected source APIs, and choosing Anthropic or OpenAI for optional theme labels sends a small sample of mention text to that provider.github.com · read 7 Oct 2026
- Purpose
- Harken tracks mentions of keywords, brands, or products across online sources and presents sentiment and themes in a local dashboard.github.com · read 7 Oct 2026
- Sentiment analysis
- Yesgithub.com · read 21 Sep 2026
- Source requirements
- Hacker News and Bluesky work without credentials; Reddit, Mastodon, RSS, X, and YouTube require credentials, configured feeds, or tokens as applicable.github.com · read 7 Oct 2026
- Sources
- Sources include Hacker News, Reddit, Mastodon, Bluesky, Stack Overflow, RSS, X, and YouTube.github.com · read 7 Oct 2026
- Storage and telemetry
- Harken stores its database locally in SQLite and says it sends no Harken telemetry.github.com · read 7 Oct 2026
- Support
- The project invites issues and pull requests, and says paid tools offer broader hosted operations and vendor support.github.com · read 7 Oct 2026
