Tidepool / site
llms.txt1 starsChanged 8 days agoApache-2.0View on GitHub
Tidepool is a zero-dependency IDE for software developers, written in Rust from the ground up by Joseph Abbott III, with no third-party crates: its own OS bindings, rasterizer, font renderer, UI toolkit, text engine, parsers, protocols and terminal emulator. https://jabbott.dev/Tidepool/ shows its gate and two example programs of its libraries, run on each commit of main that passes CI, its architecture, and its features with how far each has come. cargo xtask ci, which CI runs on Linux, macOS and…
In this repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- add-ffi-bindingSkill
- add-managed-toolSkill
- evaluate-dependencySkill
- next-taskSkill
- record-decisionSkill
- verifySkill
The file
Files
# Tidepool
> Tidepool is a zero-dependency IDE for software developers, written in Rust from the ground up by Joseph Abbott III, with no third-party crates: its own OS bindings, rasterizer, font renderer, UI toolkit, text engine, parsers, protocols and terminal emulator. https://jabbott.dev/Tidepool/ shows its gate and two example programs of its libraries, run on each commit of `main` that passes CI, its architecture, and its features with how far each has come.
Key facts:
- Status: in development, at version 0.1, and still headless: nothing opens a window yet. {{build_line}}
- Language: Rust {{toolchain}}, edition 2024; {{crates}} crates in the workspace; third-party crates: {{third_party}}
- Planned: a native GUI in jabbott.dev's design language and a terminal UI with vim-style modal editing; first-class Rust (rust-analyzer, cargo) and Markdown; managed toolchains (rust-analyzer, git, the JDK, the Android SDK and NDK, bundletool, Xcode) installed and driven through their official channels; Android and iOS builds, signed and verified; AI through language agents speaking the Tidepool Agent Protocol, and offline predictive text
- Targets: Linux x86_64 (X11 and XWayland), macOS 13 and later (universal), Windows 10 1809 and later (x86_64)
- License: Apache-2.0
- Source: {{repository_url}}
- Author: Joseph Abbott III, software engineer (https://jabbott.dev/resume/)
## Pages
- [IDE demo](https://jabbott.dev/Tidepool/): the gate's run, the two examples' output, this build, the roadmap's milestones, two architecture diagrams and every feature
- [Repository]({{repository_url}}): the source, with the design in docs/
## The gate
`cargo xtask ci`, which CI runs on Linux, macOS and Windows for every pull request and every push to `main`; this page is published only from a commit of `main` that passed it on all three. Here: {{gate_line}}. Its summary:
```
{{gate_summary}}
```
## Formats
What `cargo run -p tidepool-codec --example formats` printed:
```
{{formats_output}}
```
## Text engine
What `cargo run -p tidepool-text --example editing` printed:
```
{{editing_output}}
```
## Architecture
The page draws two views, which {{repository_url}}/blob/{{commit}}/diagrams/c4-architecture.md also draws, with every module and edge:
- Containers: the developer is to use Tidepool through its GUI frontend (roadmap M2.6) or its terminal UI (M7.1), both parts of the tidepool binary, whose main is empty today (M0.1). The binary is to read and write its settings and other TOML files (M3.4, M4.7, M5.1), open, edit and save the developer's workspace folders (M4.7, M4.3), install and run managed tools such as curl, rustup and cargo, rust-analyzer, git, the Android SDK and Xcode as child processes (from M5.1), and speak the Tidepool Agent Protocol over stdio to language agents (M9.3); the reference agents reach model servers through curl or loopback HTTP (M9.6). The core binary opens no sockets.
- Crates by layer: 26 crates in seven layers, each using only crates below it that its row of the dependency matrix names. Built: core (M0.2, M1.6, M1.8) and sys, its crate root (M0.1) and Linux declarations (M2.1), in layer 0; codec (M1.4 to M1.7), text (M1.1 to M1.3) and the crate root of platform (M0.1) in layer 1; the binary with an empty main in layer 5 (M0.1); xtask (from M0.3) and bench (M1.9) in layer 6. Only sys and platform may hold unsafe code.
## Features
{{features_lede}}
{{features}}
## Roadmap milestones
{{milestones}}
## Optional
- [Roadmap]({{repository_url}}/blob/{{commit}}/docs/roadmap.md): every task, with its acceptance criteria
- [Architecture]({{repository_url}}/blob/{{commit}}/docs/01-architecture.md): the crates, their layers and dependency matrix
- [Decisions]({{repository_url}}/blob/{{commit}}/docs/decisions.md): the architecture decision records
- [Diagrams]({{repository_url}}/tree/{{commit}}/diagrams): C4 views of the system, its programs and its crates, components, classes, sequences, data flow and states
- [jabbott.dev](https://jabbott.dev/llms.txt): the author's site: resume, engineering log and other projects
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

