AI Coding Engine · Built for Perplexity
Plan Mode, persistent memory, and a full Claude Code workflow — all inside a Perplexity Space. Four prompt files. One Space. Start building.
System Active
// Four engines
NOT A PROMPT. A SYSTEM.
v3 splits GAIA Code into four prompt files that drive four engines — so it remembers, plans, paces itself, and runs your skills.
Survives auto-compaction.
GAIA keeps a MEMORY.md in its sandbox — project structure, your standing notes, and observations it records itself. It re-reads memory after every compaction and writes down mistakes it fixed so it never repeats them.
Explore. Plan. Approve. Execute.
Non-trivial work starts as a written PLAN.md you approve before a line is touched. Execution is tracked in TASKS.md with checkboxes flipped the moment each task lands — never batched at the end.
Never overflows the window.
Before generating, GAIA estimates the token cost of a turn and stops at a safe boundary, so Perplexity doesn't crash mid-task. Commits are batched by file size for clean, reliable history.
Slash-command skills.
Drop a skill file into the Space and call it with a slash command (e.g. /humanizer). GAIA adapts external, non-ported skills to the tools it actually has.
Read repos, create branches, push commits, open PRs, and manage issues — all in one Perplexity conversation.
// Architecture
ONE GATE. THREE ENGINES.
Paste SYSTEM_INSTRUCTIONS.md into the Space, upload the three engine files, and GAIA wires itself up. See exactly how it fits together.
SEE THE ARCHITECTURERegistry-pinned dependencies
Versions come from the package registry's JSON API in real time — never a stale search snippet. No deprecated or vulnerable pins.
Self-review before every push
The sandbox can't compile, so GAIA re-reads generated code for the errors a build would catch before it ever pushes.
Four files in, GAIA Code handles the rest — memory, planning, and clean commits.