[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:operator-memory":3},"\u003Cp>\n  \u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Faerovato\u002Foperator-memory\u002FHEAD\u002Fdocs\u002Fassets\u002Fbanner.jpeg\" alt=\"Operator Memory, the self-improving context engine for coding agents\" \u002F>\n\u003C\u002Fp>\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002FREADME.md\" rel=\"nofollow ugc noopener\">English\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002FREADME.zh-CN.md\" rel=\"nofollow ugc noopener\">简体中文\u003C\u002Fa>\n\u003C\u002Fp>\u003Chr \u002F>\n\u003Ch3>Operator Memory: The self-improving context engine for coding agents.\u003C\u002Fh3>\n\u003Cp>Operator Memory gives your agent a brain for documenting all their work. As the agent works, it automatically documents within this brain — specs, decisions, standards, research, lessons. Every new session starts knowing everything the last one learned.\u003C\u002Fp>\n\u003Ch4>Features\u003C\u002Fh4>\n\u003Cul>\n\u003Cli>\u003Cstrong>Complete Context Engine\u003C\u002Fstrong> — Documentation, memory, indexes, and skills within a single integrated system.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Automatic Documentation\u003C\u002Fstrong> — Specs, decisions, research, and more are automatically documented by the agent.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Transparent Memory\u003C\u002Fstrong> — Memory is stored as Markdown documents you can read, update, and delete.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Sharable Knowledge\u003C\u002Fstrong> — Track documents with Git and share knowledge with your team.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Zero Infrastructure\u003C\u002Fstrong> — No background agents, no embeddings, no vector database, no model configuration.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3>Installation\u003C\u002Fh3>\n\u003Cp>Operator is managed via the Operator Helper, which also provides setup and agent instructions. Install the Helper with npm:\u003C\u002Fp>\n\u003Cpre>\u003Ccode class=\"language-sh\">npm install --global @aerovato\u002Foperator-helper\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>Or Bun:\u003C\u002Fp>\n\u003Cpre>\u003Ccode class=\"language-sh\">bun add --global --minimum-release-age 0 @aerovato\u002Foperator-helper@latest\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>Then install the Operator adapter for your harness. Click each link for harness-specific information.\u003C\u002Fp>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Harness\u003C\u002Fth>\n\u003Cth>Status\u003C\u002Fth>\n\u003Cth>Install\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fclaude-code.md\" rel=\"nofollow ugc noopener\">Claude Code (CLI + Desktop)\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟢 Fully Supported\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install claude-code\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fcodex.md\" rel=\"nofollow ugc noopener\">Codex (CLI + Desktop)\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟢 Fully Supported\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install codex\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fopencode-v2.md\" rel=\"nofollow ugc noopener\">OpenCode\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟢 Fully Supported (V2)\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install opencode\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fopencode.md\" rel=\"nofollow ugc noopener\">OpenCode V1 (legacy)\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟡 Supported, Legacy\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install opencode-v1\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fpi.md\" rel=\"nofollow ugc noopener\">Pi\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟢 Fully Supported\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install pi\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fdeepseek.md\" rel=\"nofollow ugc noopener\">DeepSeek Harness\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟢 Fully Supported\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install deepseek\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fharnesses\u002Fkiro.md\" rel=\"nofollow ugc noopener\">Kiro\u003C\u002Fa>\u003C\u002Ftd>\n\u003Ctd>🟡 Supported\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>operator-helper install kiro\u003C\u002Fcode>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Cp>Setup is a conversation with your agent. Run each command in a new conversation.\u003C\u002Fp>\n\u003Col>\n\u003Cli>First time only: \u003Ccode>\u002Foperator:user-init\u003C\u002Fcode> — set up your global user partition.\u003C\u002Fli>\n\u003Cli>In each new project: \u003Ccode>\u002Foperator:project-init\u003C\u002Fcode> — scaffold Operator, migrate existing documents, and index the repo.\u003C\u002Fli>\n\u003Cli>Start a new conversation and do normal work.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Cp>When starting cold on an existing project, expect a sparse brain at first; ask the agent to write the first specifications for the modules you work on. Once those documents exist, later sessions maintain them as part of ordinary work.\u003C\u002Fp>\n\u003Cp align=\"center\">\u003Cstrong>New to Operator? Read the \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Faerovato\u002Foperator-memory\u002Fblob\u002FHEAD\u002Fdocs\u002Fgetting-started.md\" rel=\"nofollow ugc noopener\">Getting Started guide\u003C\u002Fa>.\u003C\u002Fstrong>\u003C\u002Fp>\u003Ch3>How Operator Works\u003C\u002Fh3>\n\u003Cp>Agents excel in a single session but forget everything the moment it ends. Future sessions waste tokens re-gathering an incomplete context: re-exploring the codebase, re-learning the architecture, re-teaching decisions and corrections.\u003C\u002Fp>\n\u003Cp>Operator gives the agent a durable workspace of Markdown, kept in three places:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Ccode>.operator\u002F\u003C\u002Fcode> — private project knowledge, stays on your machine\u003C\u002Fli>\n\u003Cli>\u003Ccode>.operator-shared\u002F\u003C\u002Fcode> — project knowledge published with the repository\u003C\u002Fli>\n\u003Cli>\u003Ccode>~\u002F.operator\u002Fuser\u002F\u003C\u002Fcode> — your personal rules and knowledge, used across projects\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Every session runs the same loop:\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>Consult\u003C\u002Fstrong> — the agent starts from your Brain: instructions, codebase index, specs, guides.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Build\u003C\u002Fstrong> — the agent does normal development work, informed by that knowledge.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Update\u003C\u002Fstrong> — the agent records what changed: new specs, decisions, standards, lessons.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Cp>When project truth changes, the agent updates the canonical file instead of adding a RAG database record. For more details, see [Architecture](docs\u002Farchitect\u003C\u002Fp>\n",1793926097144]