Long-term memory
for every agent.
Forget keeps what was said, word for word, and rebuilds what an agent needs at the moment it asks. Every line cites its source. When the record is silent, it says so. One memory for Claude Code, Codex, Claude Desktop and any MCP agent.
Don’t rewrite memory.
Rebuild it.
Most memory layers summarize or extract facts when they save. Whatever the summary drops is gone, and the answer loses its source. Forget never rewrites the record. It does the work when a question arrives.
One memory,
three moments.
Forget works when a message is saved, while your agents are idle, and when a question comes in. Every derived item points back to the verbatim record, so any answer can be checked against what was actually said.
↻ Real usage teaches the memory model what to keep, link and recall.
Decide once.
Every tool knows.
A decision made in Claude Code on Tuesday is there when Codex opens on Wednesday. When your server agent runs that night, the version it replaced is flagged before it can be used.
An example project in a simplified terminal. Every step is a real Forget tool call.
Memory you can check.
Decisions, not transcripts.
An observation gate keeps the facts that matter and lets the chatter go. Decisions also surface from your git history, backdated to the commit.
A receipt on every answer.
Every memory links to the commit or conversation that decided it, with a trust label: green if you said it or a tool saw it, yellow if an agent inferred it, red once it is superseded.
Stale facts get caught.
A changed fact is superseded, not deleted. An agent about to act on the old version sees both, labeled, before it acts.
Every session opens briefed.
Start Claude Code or Codex and a capsule is waiting: the current goal, the next step and anything that has gone stale.
Recall, measured.
How often the right memory lands in the top six results, on our hardest benchmark: 120 questions, six question types, multi-part answers scored by coverage.
Every number was measured before its engine shipped. Earlier end-to-end result: 78.4% ± 0.4 on LongMemEval, three seeds.
Measured in public.
On the Agent Memory Leaderboard, an independent benchmark where a fixed reader answers only from what the memory returns.
9,602 questions, text and streaming
Cycle 1 leader: 58.4
Unseen questions, vs. verbatim search alone
Cycle 2 official rankings are published in mid-November. The +4.5 result comes from our research engine; managed reconstruction is rolling out to the hosted ledger next.
Connect in a minute.
One engine, one command per machine. Using an AI agent? Point it at forget.sh/llms.txt and it can set this up for you.
Install the engine
Installs the Forget engine and connects Claude Code, Codex and Claude Desktop. Your existing MCP setup is backed up first.
Share one ledger
Run it on each machine and every client writes to the same ledger. Host it on your own server, or get early access and we host it and send you its URL and token.
Undo in one command
Restores your original client configs from backup and removes the hooks. Other settings are left exactly as they were.
Pricing.
Self-hosted
Free
- The full engine, on your machine or your own server
- Every connected client shares one ledger
- Deep recall on your own local model
Shared ledger
Early access
- We host one ledger for all your tools and machines
- One ledger file per account, hashed access tokens, daily backups
- Invite-only during early access
Cloud Pro
$8/ month
- 2,000 deep recalls a month on the certified model
- About 2.5 s answers, no GPU needed
- Recall requests pass through and are not stored
Questions.
What exactly gets shared?
Memories: decisions, facts and task state, each with its source. Every client connected to the same ledger reads and writes it. Which tool wrote a memory is kept as provenance, not as a wall.
Where does my memory live?
Wherever the ledger runs. Self-hosted, it is a SQLite file on your machine or server. With the shared ledger (early access), we host it: one ledger file per account, hashed access tokens and daily backups. End-to-end encrypted sync is in design and not shipped yet.
How is this different from ChatGPT or Claude memory?
Built-in memory lives inside one vendor’s app and keeps summaries without sources. Forget keeps the verbatim record, is shared by every tool that speaks MCP, and every answer carries a receipt: the commit or conversation that decided it.
What is query-time reconstruction?
Instead of compressing conversations into facts when they are saved, Forget keeps them word for word and assembles the evidence a question needs when it is asked. Each sentence of that evidence cites the raw record it came from, and if nothing supports an answer, the agent is told to abstain.
What happens when a fact changes?
The old fact is not deleted. It is superseded: it keeps its history, turns red, and the new fact links back to it. If an agent is about to act on the stale version, the conflict alert shows both first.
What does it cost?
The engine is free. Cloud Pro, for deep recall without your own GPU, is $8 a month. The hosted shared ledger is in early access.
How do I remove it?
Run npx forget-connect disconnect to restore your original client configs and remove the hooks. Self-hosted, delete the ledger file. For a hosted ledger, write to hello@forget.sh and we delete its file.
What is supported today?
macOS and Linux. Claude Code, Codex and Claude Desktop through forget-connect, and any MCP client over HTTP. Not shipped yet: Windows and end-to-end encrypted sync.
Recently shipped.
- Oct 19
Hosted ledger moves to the 0.6 engine.
Every message is now kept verbatim alongside its facts, search combines BM25 with embeddings, and each result reports whether the evidence is strong, weak or missing.
- Sep 24
Shared ledger in production.
Claude Code, Codex and server agents can now share one hosted ledger: one ledger file per user, hashed per-user tokens, daily backups.
- Sep 5
0.5.0, the honesty release.
An unverified fact with a date or amount now asks for confirmation before use, machine-transcribed facts wait in quarantine, and search weighs sources. Learned forgetting and nightly consolidation.
- Aug 4
The recall dial and Cloud Pro.
Four measured gears on one dial, and deep recall on the certified model for $8 a month.
Share one memory across everything.
We host your ledger and help you connect your tools to it.
Thank you.
We will email you as soon as your spot opens.