Assertion Memory
Every Claude Code, Cursor and Codex session builds on what your team already learned: the decisions, what was tried, and why. Nobody re‑explains, and nobody repeats an experiment a teammate already ran.
Works with Claude Code, Cursor and Codex. Prefer one app? Our agent: no compaction, half the price →
15,988 countingdecisions and facts remembered so far
Who uses it?
“Our team works from one shared memory. When someone picks up our algorithm work, their coding agent already knows what everyone else has tried and why, so we build on each other's work instead of repeating it. Day to day, it's also made my LLM interactions more productive.”
“My architecture work needs to consider hundreds of components over long timelines. With Assertion, I spend less time assembling and maintaining context. It gives each new Claude Code session fast access to my whole history of decisions, without blowing the token budget.”
From a customer’s memory
A real sequence from a customer’s memory. The dates and counts are unchanged; the subject matter has been changed.
How does it work?
While your agent works, memory keeps the decisions and their reasons: what was tried, what failed, and what was settled. Not your transcripts or source code.
A new session starts from what earlier sessions learned, even weeks later, without anyone digging up old chats or pasting in context.
Put teammates in a shared space and each person’s agent recalls what the others captured there. Memory stays private to you until you share a space.
On BEAM 100K, a public benchmark for long-term memory, Assertion Memory scores 81.1%, ahead of the leading memory system. See the benchmark →
Can we trust it?
Prefer one app?
The Assertion agent is our own coding agent for Mac. Memory holds the session, so it works the way the plugin does, and it needs far fewer tokens, so it costs less.
Memory keeps what matters as you work, so long sessions don’t turn lossy. In a 2026 study, compaction kept only 17% of the instructions given during a session. Wang et al.
3×+ fewer tokens on long sessions, passed on: pay $10, get $20 of model usage.
Claude, GPT, Gemini, Grok and Muse, on one bill. For Mac with Apple Silicon.
How do I start?
Add the free plugin to the agent you use, or use Assertion’s own agent.
Teams, per memberThe same $10 Team seat as the agent, shared memory included. 14-day trial, no seat minimum.
For Claude Code, Cursor and Codex, on any platform they run on.
Agent plans usually cost $20, $60, $100 and $200 a month.
Teams, per member, usage pooledTeam $10 monthly or $8 annually, $20 of model usage. Team Premium $50 monthly or $40 annually, $100 of model usage, with admin controls. One seat covers a teammate on the agent or the plugin.
For Mac with Apple Silicon, or in a terminal: curl -sSL https://storage.googleapis.com/assertion-downloads/install.sh | sh
The plugin stays free for individual developers, including at work, for anyone who would rather keep the agent they have. Model usage is counted at each provider’s standard API price. Annual billing saves 20% on every paid plan. Agent tiers differ by included usage, not capability. One Team seat covers agent and plugin users alike. Assertion Analytics is licensed annually and includes Memory.
Questions
Claude Code, Cursor and Codex, through the free plugin, wherever they run. Or use our own agent, on Macs with Apple Silicon.
Only if you share a space with them. Memory is private to you by default. In a shared space, each teammate’s agent recalls what the others captured there.
On long coding sessions, at least 3× fewer than the same model on its own, memory included. The longer you work, the bigger the saving; short tasks save less.
Yes. Everything the agent remembers is in Studio, with the evidence it came from, and you can delete anything you don’t want kept. It stays private to you unless you share a space with your team.
Claude (Sonnet 5, Opus 5.5, Opus 5, Fable 5.1), GPT (GPT-6 Astra and Sol, GPT-5.6), Gemini (3.8 Flash, 3.1 Pro), Grok 4.7 and Meta’s Muse Wonder 1.0. New frontier models are added as their providers release them.
Your included usage resets every week. If you run out before then, you can upgrade at any time, or carry on after the reset. Nothing you have done is lost either way.
Not in the agent today. It runs every model through your plan, so there is one bill and memory stays in step with the work. To use your own key or subscription, add the free plugin to Claude Code, Cursor or Codex and get the same memory there.
The agent runs on Macs with Apple Silicon, as a desktop app or in the terminal. On another platform, the plugin brings the same memory to Claude Code, Cursor or Codex wherever they run.
Analytics is our second product, for business-data analysis rather than software work. It is built on the same memory and includes it. If that is what you need, start there instead.
No. Your code, data and memory are never used to train models.
Get started
Free for individual developers. Install it in Claude Code, Cursor or Codex, and share a space when your team is ready.