Start with context
Every task begins with what your team already learned. No archaeology. No repeated mistakes.
fluws context --task "..."fluws turns every hard-won engineering decision into context your team's AI agents can find, trust, and use in their next session.
Agents are great at solving the problem in front of them. They are terrible at remembering why your team solved the same problem last month.
Why is this endpoint built this way?
That answer is somewhere in a lost chat, an old PR, or someone's local notes.One small habit. A compounding advantage for every person and every agent on your team.
Every task begins with what your team already learned. No archaeology. No repeated mistakes.
fluws context --task "..."When an agent discovers the why behind the code, it leaves a durable note in plain Markdown.
fluws write --name api/decision-x --body "..."See who changed what, follow the history, and review agent contributions before they land.
fluws revalidate api/decision-xCreate a key per agent, store it in the local keystore, and run fluws init. Every project, every agent, and every teammate gets the same source of truth.
fluws keys set claude --key flw_live_•••••••• fluws init --agent claude
Search follows the links between decisions, gotchas, and specs, not just matching words.
Every document carries its provenance, version history, and the person or agent behind each change.
Your memory is Markdown. Store it with fluws, in S3, or in your own GitHub repository.
Free while we learn what teams need. No credit card. No lock-in. Just the context your agents should have had from the beginning.
Start for free→