--- title: "My Workflows — Dr. Tali Režun" url: "https://talirezun.com/workflows" description: "How Tali Režun works with AI agents: the context window, orchestration, working across tools and computers, and memory." updated: "2026-10-21T18:38:43.033Z" --- Workflows · for students and engineering leads # My Workflows I build with AI agents every day, and what decides whether it works is not the prompt. It is what each agent sees, who delegates to whom, and where the project's memory lives. Four short chapters, each with something to play with. 01 ## The context window Before the first prompt, your tool has already filled part of the window. What you add next decides whether the agent starts informed or starts guessing. Switch the layers on and off and watch what the window is made of. Playground · a session start ### What your agent sees when a session starts. Three layers fill the window before the first prompt: documents, memory and knowledge. Your tool takes its own share. Play the two approaches, switch items on and off, or change the window size. Window Your context window at the start of a session · to scale ≈79.8k in use · 40% System prompt and tools 50k free 120.2k 0 100k 200k tokens The curated part, enlarged to its own scale · ≈29.8k · 14.9% of the window Build plan 3.5k Architecture 6k CLAUDE.md · AGENTS.md 2.5k Second brain 9k Shared brain 6k Documents · foundations Memory · working state Knowledge · second and shared brain Your tool's share Uncurated 1 Documents Foundations: the canonical documents the project is built against 2 Memory Working state: what carries from one session to the next 3 Knowledge Compounded knowledge, pulled by relevance 4 Your tool's share System prompt and tools of the harness. Always there × Without curation What gets pasted in when nobody curates **79.8k** tokens in the window **40%** of the window used **100%** of what you added is curated **0.9×** relative cost per call Curated and lean The agent reads the build plan, the standing brief and the latest handoff first. It knows the goal, the rules and where the last session stopped. Illustrative numbers, close to a real project in The Curator. The tool's share is an estimate: no tool reports it. [See The Curator →](https://talirezun.com/projects#p-curator) Tools used [The Curator](https://talirezun.com/projects#p-curator) Go deeper - [Context is the Code: The Complete Three-Phase Process for Building with AI Agents](https://talirezun.com/library/context-is-the-code) - [From Prompts to Precision: The Art & Science of Context Engineering](https://talirezun.com/library/art-and-science-of-context-engineering) - [The Right Context, Not All of It](https://talirezun.com/library/right-context-not-all-of-it) 02 ## Orchestration My main session plans, delegates, syncs state and audits. It does not do the bulk work. Each task goes to the cheapest model that can do it well, and nothing is accepted until it has been checked. Step 0 of 9 1. Rename files to the new convention Queued 2. Build the Library page Queued 3. Design the entitlement schema Queued 4. Audit a diff Queued Orchestrator Plans, routes, syncs state, audits. Reads summaries, not workings. Orchestrator context 12% Current worker context 0% A worker can fill its window; it is thrown away after the audit. - Small model Simple, mechanical work today: Haiku - Mid model Front-end work today: Sonnet - Top model Complex back-end work today: Opus Four tasks are queued. In orchestrate mode each one is routed by complexity; in one long session the same agent does all of them. Today: Haiku / Sonnet / Opus. The tiers are the idea; the model names change every few months. Illustrative percentages of one context window, to show the shape, not to measure a tool. Tools used [The Curator](https://talirezun.com/projects#p-curator) [Conduit and Atlas](https://talirezun.com/projects#p-conduit) Go deeper - [From One Agent to Coding Agent Armies: My 15-Month Journey to AI Orchestration](https://talirezun.com/library/coding-agent-armies-orchestration) - [Context is the Code: The Complete Three-Phase Process for Building with AI Agents](https://talirezun.com/library/context-is-the-code) - [The Mixed Fleet](https://talirezun.com/library/the-mixed-fleet) 03 ## Multi-harness work I move between coding tools and between computers on the same project. That only works if the project's state does not belong to any one tool. Step through a session that changes tool and machine and still picks up where it stopped. Step 1 of 7 Machine A **Claude Code** scope: claude-code **opencode** scope: opencode Machine B **Antigravity** scope: antigravity A Curator project · a folder you own plain markdown - **Foundations** canonical documents reading - **Standing brief** your standing orders reading - **Handoff** scope: claude-code not saved yet - **Handoff** scope: antigravity not saved yet - **Handoff** scope: opencode not saved yet - **Journal** recent sessions reading **Session starts.** Claude Code opens the project and makes one call to read its context: foundations, the standing brief and the latest handoff. All of this is plain markdown in a folder you own. It is portable, and it is not locked to any one vendor, tool or computer. Tools used [The Curator](https://talirezun.com/projects#p-curator) [AgentLink](https://talirezun.com/projects#p-agentlink) Go deeper - [The Handoff Writes Itself](https://talirezun.com/library/the-handoff-writes-itself) - [Where Your Context Lives](https://talirezun.com/library/where-your-context-lives) - [The Mixed Fleet](https://talirezun.com/library/the-mixed-fleet) 04 ## Memory that compounds Not all context is the same, so it should not be written the same way. Documents are kept verbatim, working state is replaced at each save, and knowledge accumulates. Select a tier to see what it holds and who writes it. Project context What an agent reads at the start of a session Knowledge Compounded, pulled by relevance Agents read this at session start and write it as they work. Foundations Holds The documents the project is built against: plan, architecture, decisions. Some are read first, some on request. Written by You, or an agent you commission. A model never rewrites one. Changes Rarely. When one changes, the document is replaced whole. Rule Read verbatim Knowledge accumulates. State supersedes. A canonical document is replaced whole and read verbatim. All of it is plain markdown on your own computer, read and written by agents through The Curator's built-in MCP server. It runs on your own API key, from one of three providers. [The Curator at mycurator.xyz](https://mycurator.xyz) [⭐︎ Star The Curator on GitHub](https://github.com/talirezun/the-curator) Tools used [The Curator](https://talirezun.com/projects#p-curator) Go deeper - [The Agent Memory Problem — And Why Your Second Brain Might Be the Answer](https://talirezun.com/library/agent-memory-problem) - [Where Your Context Lives](https://talirezun.com/library/where-your-context-lives) - [The Context Engine](https://talirezun.com/library/the-context-engine) --- HTML page: https://talirezun.com/workflows · Site index for agents: https://talirezun.com/llms.txt