Application Stack · search
SearXNG
Run a private search front end for personal or team use instead of relying on one commercial search provider interface.
Workflows the agent can manage
- search the web
- aggregate multiple search engines
- configure search sources
- operate a private search endpoint
Agent operability
Editorial agent-readiness estimate: 4/5. This reflects documented interfaces and proposed workflows, not a successful installation or a support promise. Assessment date not recorded; verify upstream documentation before use.
- REST — Search responses can be exposed in machine-readable formats when enabled in SearXNG configuration
Evidence consulted
- Official README describes SearXNG as a metasearch engine whose users are neither tracked nor profiled and links current installation/configuration guidance. [stated; https://github.com/searxng/searxng/blob/master/README.rst]
- Official package metadata identifies the licence as AGPL-3.0-or-later. [stated; https://github.com/searxng/searxng/blob/master/setup.py]
Implementation notes
Deploy the maintained upstream service, configure only the engines/formats the user needs, enable rate limiting for exposed instances, and treat upstream engine availability/terms as external dependencies.
Licence / edition: AGPL-3.0-or-later Public instances require careful limiter/bot configuration and must respect upstream search-engine terms and limits.
Resources: light class — SearXNG provides maintained container images and can run as a small web service; actual resource use depends on concurrency, enabled engines and rate limiting. No trustworthy fixed RAM baseline is asserted here. (inferred)
External prerequisites: network access to enabled upstream search engines
Security: Catalogue qualification is not a security certification; inspect current advisories, supported versions and deployment guidance before setup.