reflect
Long-term memory for coding agents. Correct a mistake once, never again.
How it works
One loop, running in the background of every session. Same engine and knowledge base across Claude Code, Codex CLI and GitHub Copilot.
Install and first run
The engine is a Python package. Install it with uv, then create a knowledge base and search it.
uv tool install --upgrade --torch-backend cpu \ 'git+https://github.com/stevengonsalvez/ainb-reflect-memory.git[graph]'reflect --versionreflect initreflect search "how did we fix the tokio panic"Find your way around
Start with the overview, then pick the harness you use.
Start hereWhat reflect is, the problem it fixes, a five minute quickstart, and how it compares.
Install per harnessClaude Code, Codex CLI, GitHub Copilot and Hermes setup.
ConceptsArchitecture, capture, drain, index and storage, and the recall pipeline.
InteractiveMemory browser, recall walkthrough, token economics and hook timeline you can poke at.
GuidesThe memory browser, the TUI plugin, fleet use, and troubleshooting.
ReferenceCLI, hooks, configuration, KB format, OKF profile and the serve API.
Evals and benchmarksRecall benchmarks and the regression suite that guards them.
Design notesWhy reflect is shaped the way it is, and how it relates to OKF.