Skip to content
Lossless
Esc
↑↓navigate↵open⌘Jpreview

Comparison

How the Pi and OpenCode v2 adapters compare with the engines and memory services their users choose between, and why the comparisons are separate.

The engine is shared. The peer sets are not.

Pi sessions pick among compaction engines. OpenCode v2 sessions pick among context engines, memory services, or both, so each adapter has its own page.

What a cell means

Each cell comes from the peer’s source at the commit named in that page’s verification note, not from a README. The date on the page covers the peer cells and this package’s own. Each page names where a peer leads and which gaps this package keeps.

The audited sources are comparison/pi.mdx and comparison/opencode.mdx. The page and the source are the same file.

Last updated on

Was this page helpful?