Amnesia
Decisions, preferences, and project state disappear when the session ends. A useful agent should carry forward what matters.
Durable memory over session-only state.
Persistent personal AI
Your own persistent AI
Aegis is your own persistent AI: who remembers, reasons, and endures as your work, preferences, and history evolve.
Long-term memoryLong-horizon decisionsLong-context understanding
Why Aegis exists
Most current agents fail in three predictable ways.
Session-only behavior does not hold up when an agent is meant to stay useful over time. Aegis exists to replace forgetting, drift, and overflow with durable recovery.
Myopia
Current-turn optimization wins over next-week usefulness. Aegis is opinionated about persistent goals, decomposition, and milestone awareness.
Long-horizon decisions over local cleverness.
Overflow
When the context window fills up, naive truncation destroys the structure that should have been remembered, compressed, or retrieved.
Layered recovery over blind truncation.
CLI-first quickstart
Install once, launch Aegis, and start talking.
The public path should be direct and truthful: one remote install command, one local CLI, and one durable operator path from `init` through `wake`.
Bootstrap the public Aegis CLI.
This is the supported install path for the public site. It prepares the local runtime, writes the `aegis` launcher, and launches Aegis immediately.
curl -fsSL https://aegis.agentic-in.ai/install.sh | bashmacOS / LinuxPython 3.12+Writes `~/.local/bin/aegis`
Then continue in docs
Do the rest of the flow once, not from the homepage.
After install, stay inside the CLI-first path: run `aegis init`, confirm with `aegis status`, create a clone if you need another continuity line, and re-enter it through `wake`.
Core product thesis
The system is organized around three durable pillars.
These are the claims the project is built to make legible: long-term memory, long-horizon decision making, and long-context understanding.
Long-term memory
Durable episodic, semantic, and procedural memory instead of session-only state.
Long-horizon decision making
Persistent goals, task decomposition, and deadline-aware planning across time.
Long-context understanding
Layered context management, compression, and retrieval instead of naive truncation.
What Aegis points toward
A personal AI that stays yours across years, not sessions.
The aim is an agent that becomes more useful the longer it knows your work: it remembers what matters, reasons over long horizons, and returns with the right context intact.