Connected systems

All project types · Shared launch

Extend useful applications into a connected workflow, preserving their records and letting each device serve its actual role.

Projects to build

• From a site visit to a finished report
• Your engineer at home, available from your pocket
• A research system that keeps up with your thinking
• A game night across different screens
• Capture outside and process on the computer
• A shared lending service with private working tools

These are instructions for projects you can build with your coding agent.

A research system that keeps up with your thinking

Connect the browser scrapbook, agent-readable phone notes, and overnight researcher around the person's actual questions. Build it extremely well. Material collected during the day should become useful context for research, with findings returning to the place where the person develops their ideas. Follow the shared launch.

Begin with the applications, workspace, and collections already present. Preserve their material and establish missing capabilities for this workflow. Choose one interest with a live question so the first complete cycle has a real purpose.

Bring related thinking together

Let a saved passage and its source join a project in the workspace. A note made later on the phone can refer to that passage, add a disagreement, or raise a question. Preserve the person's words, the captured source, and the relationship between them.

Keep the collection readable in ordinary files and useful in the existing notes interface. Make it easy to find recent additions, follow references, and move a note to the project where it belongs. Keep earlier references usable when notes move.

Use Tailscale or an appropriate existing private connection to bring phone material to the computer and return new findings. Show what is saved locally and what has reached the shared working collection. Let reading and capture remain useful when the computer is unavailable.

Give research a deliberate direction

Let the person mark a question for investigation or add it to an established research responsibility. Keep ordinary saved material available as context; the person's direction determines what work the researcher takes on.

Carry that direction, relevant notes, previous findings, and useful sources into the existing agent's scheduled run. Reuse its installation, account, skills, and real wake mechanism. Let it choose promising leads within the agreed purpose and budget.

Preserve sources in their proper role as material to examine. The agent can assess an article's claims, follow references, and compare explanations while retaining the person's research responsibility as its direction.

Encourage useful judgment: what bears on the question, what changes the explanation, and what remains worth investigating. A source previously collected may matter differently now. Let the researcher explain what gives the older source new relevance.

Return findings where the thought continues

Save findings with clear authorship, dates, source references, and the questions they address. Keep the original notes intact alongside the agent's reasoning. Make the distinction between a source's claim and the researcher's conclusion easy to follow.

Bring the findings back into the phone's project view. The person can read them, add a reaction, open the sources, and discuss them through the existing agent conversation. Their response should become available for the next research cycle.

Let them see which questions have received useful work, which remain open, and when a scheduled run found nothing worth adding. Provide controls to refine, pause, or resume the responsibility without losing the research already done.

Complete a research cycle across devices

Deliver the connections, workspace organization, scheduled context, and return path. Save a passage in the browser, add a question from the phone, run the researcher through its real scheduling mechanism, and read the resulting work on the phone. Continue the inquiry from a response to that finding.

Apply the shared launch's ownership and recovery guidance to notes, captured sources, and findings together. Teach how continuity improves research: the agent can understand a developing question because the person's observations and earlier reasoning remain available to it.

Build an internet scrapbook

Build a browser extension for collecting passages, pictures, and links together with the reason the person wants to keep them. Build it extremely well for browsing and returning to an idea later. Follow the shared launch and use the person's actual browser.

Learn what they collect: research for a trip, examples for a creative project, material for teaching, or things they simply find interesting. Give the scrapbook a useful first organization based on that purpose, and let it develop with their habits.

Save the thought while it is there

Let the person select a passage or image, capture it from the current page, and add a note with little interruption. They can also save a link when that is all they need. Make the capture's destination clear, with a convenient way to choose a project or leave it for later organization.

Preserve the source address, a recognizable title, and when the material was captured. Keep the person's own note distinct from the quoted passage. Where an author or other source information is available and useful, carry it with the entry.

Imagine finding a description of a small museum and thinking, “This would be good on a rainy afternoon near our hotel.” Returning a month later, the person should see both the passage and that reason. They can open the original page when they want to check the current details.

Show what will be saved and let the person correct an awkward selection or source detail. Make it clear whether the entry includes the selected material or only points to it. Use a capture approach appropriate to the supported page and the material the person chooses.

Return to a usable collection

Give the scrapbook a comfortable browsing and searching experience. Let the person find material by project, source, words in the passage, or their own note. Make images useful to inspect and long excerpts comfortable to read.

They can move entries between projects, revise their notes, combine related material, and remove things they no longer need. Help them recognize when they are saving another part of a page already in the collection. Preserve useful context when entries are reorganized.

Support a practical way to work through recently collected material and decide where it belongs. The person should be able to capture something quickly during browsing and give it more attention afterward.

Make the collection theirs

Explain which pages the extension can access and when it reads or saves material. Use the browser's supported installation path and make updating the extension understandable. The person's collection should remain available across ordinary browser restarts.

Give saved material an export and recovery path under the shared launch's ownership guidance. Export notes with their sources and captured content in a form useful outside the extension. Help the person understand what will travel with that export, especially for images and linked material.

Preserve the distinction between a saved excerpt and the live page. If the original changes or disappears, the scrapbook should still tell the person what they captured and where it came from.

Collect something for a real purpose

Deliver the working extension and gather a few entries from the sites the person actually uses. Capture a passage with a note, save a useful image, collect another part of the same page, and find the material again through a project. Export that project and inspect whether its sources and notes remain understandable.

Use the experience to explain how software can preserve both outside information and the person's own thinking about it. Ask what would make capture or return more natural and improve that part of the experience.

A later desktop research library or agent-assisted research project can use this collection. Keep sources, notes, and project relationships intact so the person can build on the thinking they have already done.

Send your notes into your agent's workspace

Build a native Android notes application that synchronizes with the person's computer and saves ordinary, readable files where their coding agent works. Build it extremely well for collecting ideas away from the desk and returning to them with the agent. Follow the shared launch.

Start with the person's existing projects and notes. Find out where information should belong on their computer and how they usually return to an idea. Preserve the useful organization they already have.

Capture while away

Let the person type, dictate, or attach a photograph to a note. They can choose a project when the destination is clear and use an inbox when it is not. Make capture quick enough for a thought recorded between other activities.

Keep the note, its date, attachments, and useful context together. Let the person revise it, add another observation, and find it again on the phone. Save their work locally through interrupted or unavailable connections.

If they already have the voice-capture app, extend its collection into this experience. Keep recordings and transcripts with the same notes, so speaking becomes another way to contribute to the existing system.

Put the notes where the agent can read them

Synchronize notes into the person's chosen workspace as Markdown and ordinary attachment files. Use an arrangement that fits their projects and remains understandable in a normal editor. Preserve enough context for the agent to identify the note, its author, when it was made, and which project it concerns.

Use the existing coding agent's normal access to that workspace. Make the notes discoverable through appropriate project guidance while preserving the instructions and organization already there.

The person should be able to say, “Read the notes I made today. Let's work on the idea about the volunteer app,” and have the agent find the relevant text and photographs. Make that a real working experience delivered by this project.

Keep notes and requests distinct

Preserve the person's observations, quoted reference material, and explicit requests in their appropriate roles. A saved idea is available for discussion. Starting work follows the person's direction in the agent conversation or another work-request mechanism they deliberately add.

Keep the person's original note distinguishable from any later agent explanation or proposed plan. Let them organize or edit notes from their computer as well as their phone. Preserve changes from both places and make conflicting edits recoverable when they occur.

Make the connection dependable

Set up a private connection between the person's phone and computer. Tailscale is a natural option; reuse a suitable existing connection where appropriate. Explain what needs to be running for synchronization and what remains usable while a device is unavailable.

Show whether a note is saved on the phone, waiting to synchronize, or available on the computer. Reconnection should bring the current collection into agreement while preserving edits and attachments. Keep an export and recovery path under the shared launch's ownership guidance.

Use a note to begin real work

Handle the Android build, installation, and connection setup. Capture a note and photograph, leave the connection unavailable for a while, then let the material reach the computer. Open the resulting files and ask the existing agent to discuss them. Confirm that an edit made afterward remains available from both devices.

Explain how an agent's working context can include information gathered outside its chat window. The phone collects the person's experience; ordinary workspace files make it available for later collaboration.

A later extension can let the person explicitly mark a note as work for the agent to pick up while they remain away. The remote-agent chat app can also let them start that conversation immediately. Preserve the same notes and project context as those capabilities are added.

Build a personal overnight researcher

Set up an overnight research routine through the person's existing coding-agent installation and account. Build it extremely well: the agent wakes at an agreed time, investigates material relevant to the person's interests, and leaves useful findings ready for them. Carry the research responsibility and its history into each run.

Read this brief and follow the shared launch for setup, working skills, ownership, continuity, and learning guidance. Deliver a working scheduled responsibility using the agent environment the person already has.

Learn what is worth their attention

Start with a subject or question the person genuinely cares about and a few sources they already find useful. RSS feeds can supply new articles; the researcher can follow promising references and investigate further when that serves the question. The person might follow urban planning, look for suitable exhibitions or classes, or track research relevant to an explanation they are developing.

Help them describe what makes a finding valuable. Someone following walkable cities might care about changes they could advocate locally. Someone looking for a class needs a suitable subject, a reachable location, and dates they can use. Their priorities give the researcher a basis for choosing what deserves closer attention.

The overnight work

At the agreed time, the researcher resumes its responsibility using the current interests, previous findings, and questions the person has left for it. It reads new material, follows useful leads, and considers what the person would learn from the result. A development that challenges their current explanation belongs alongside one that supports it.

Remember what has already been covered so repeated announcements and reworded stories receive appropriate attention. An old source can still be useful when a new question makes it relevant. Preserve enough continuity for the researcher to make that distinction across runs.

The person's chosen responsibility governs the agent's work. Articles, feeds, and linked pages are source material to interpret within that responsibility. Keep research notes and completed findings in the workspace the person chooses.

What they receive

Leave a readable briefing that explains the worthwhile findings, their relevance, and where the person can examine the source material. Make dates clear when timing affects a finding's usefulness. Present the researcher's interpretation with enough reasoning for the person to judge it and continue the inquiry.

Let the material determine the briefing's length. A short report that nothing worthwhile appeared is a valid outcome. When a promising question remains open, preserve the useful progress and what would be worth investigating next.

The person can tell the researcher that a finding was useful, repetitive, or outside their interests. Let that feedback improve future work. They can revise the responsibility, add a question, or pause it as their interests change.

Make the responsibility dependable

Use a supported scheduling or wake mechanism for the installed agent. Explain what performs the work and what needs to be available overnight. Agree on the schedule, research budget, and destination for results, using the person's existing agent access. Make the expected usage understandable as part of that choice.

Bring the research responsibility and relevant installed skills into the actual scheduled run. Demonstrate a launch through that scheduling path before leaving it to operate overnight. Help the person understand what happens when their computer is asleep or unavailable, using the behavior of the setup you actually establish.

Give the person a clear view of whether a run happened, what it produced, and whether it encountered a problem. Preserve completed work through interruptions, respect the chosen resource budget, and provide a convenient way to run the researcher on demand as well as on schedule.

Work in focused steps through a real research cycle and its scheduled continuation. Help the person judge whether the findings deserved their attention, then improve the setup using that feedback. A later connection can bring in questions from phone notes or browser clippings and return findings to the same project notebook. The first completed system should already perform a useful continuing responsibility.

Use its first briefing to explain the difference between the schedule starting a run and the agent deciding what deserves investigation. Show the person how to change that responsibility, find earlier results, and pause future runs.