ContextQMD
Back to Flow

Rise

docs/rise.md

0.1.36.0 KB
Original Source

Rise

This document explains how Rise integrates with Flow and why installing Rise gives you a high-leverage workflow layer across repos.

Rise itself is treated as closed/internal product code in many environments, but its operator model and command surface can still be documented and shared.

Start Here

Integration starts with one command:

bash
f install rise

After install:

bash
which rise
rise --help

You now have rise on your PATH and can use Rise workflows from any repo.

From Flow's install behavior (f install):

  • backend resolution is automatic (registry -> parm -> flox)
  • rise is a built-in known install target
  • Rise can be part of bootstrap tool install flows

What You Get From rise In Practice

Rise is not just one command; it is a workflow layer that adds:

  1. Repo adoption overlays for external/team repositories.
  2. Task detection and flow.toml generation/merge.
  3. JJ-native branch/bookmark workflow for clean PRs.
  4. Multi-platform compile/dev loops (web, Expo/mobile, Electron, COI).
  5. Mobile/TestFlight build observability and debugging.
  6. Schema workflows with generated TypeScript/Effect bindings.
  7. Sandbox verification in VM environments.
  8. AI and trace workflows that integrate with Flow and surrounding tooling.

Core Model: Overlay, Not Pollution

The defining Rise behavior is rise adopt.

For external repos, Rise creates a JJ overlay bookmark (rise) above main:

  • rise layer contains your local workflow files (for example flow.toml, .rise/).
  • Team-facing main remains untouched.
  • PR branches are created from main, so Rise files do not leak into PRs.

Typical flow:

bash
rise adopt https://github.com/org/repo
cd ~/code/org/repo
rise sync
jj new main -m "feat: clean PR change"

This is the main reason Rise is valuable in mixed team environments: private operator ergonomics without contaminating shared repo history.

Task Detection And Flow Integration

During adoption, Rise detects project tasks from common sources (for example package.json, Makefile, Cargo.toml, go.mod, pyproject.toml, justfile) and generates flow.toml.

You keep the generated baseline and extend it with your own project-specific tasks.

Useful commands:

bash
rise adopt .
rise adopt --force .
rise flow tasks .
rise sync
rise list

Key integration point: this makes Flow task execution (f <task>) available even in repos that did not originally ship with Flow conventions.

Development Loops

Rise provides higher-level wrappers while preserving underlying toolchains.

Common commands:

bash
rise dev
rise app
rise run dev --runner turbo
rise setup
rise verify

From Rise docs/repo behavior:

  • rise dev starts local dev paths with platform-aware behavior.
  • rise app is a fast app shortcut path.
  • rise run can delegate to task runners (Turbo/Flow patterns).
  • rise setup is project setup entrypoint.
  • rise verify is local verification flow.

Mobile/TestFlight Workflows (High Value)

Rise has dedicated mobile commands with structured observability:

bash
rise mobile validate
rise mobile preflight
rise mobile testflight
rise mobile builds
rise mobile logs

Why this matters:

  • validate catches JS bundle issues quickly.
  • preflight checks config + bundle + prebuild inspect.
  • testflight captures structured build events.
  • builds/logs provide post-failure visibility and faster debugging loops.

This reduces the "wait 10+ minutes to discover obvious build issues" pattern in Expo/EAS flows.

Schema Workflows

Rise includes schema lifecycle commands:

bash
rise schema init
rise schema status
rise schema diff
rise schema generate
rise schema push
rise schema validate

This supports a source-of-truth schema flow with generated app bindings (including TypeScript/Effect-oriented outputs in documented workflows).

Sandbox Workflows

Rise supports VM-backed sandbox execution and verification:

bash
rise sandbox
rise sandbox verify
rise verify --sandbox
rise sandbox kill
rise sandbox clean

Use this when you need stronger isolation for verification or reproduction loops.

AI + Traces + Operational Debugging

Rise docs also describe:

  • AI trace collection (rise logs)
  • build failure context for AI-assisted remediation (rise work --errors paths)
  • integration patterns around generated prompts and operational context

For Flow users, this complements:

  • f commit review/message provider strategies that can call Rise-backed providers
  • task failure hooks that invoke rise work

Typical Onboarding Sequence

For a new machine:

bash
f doctor
f auth login
f install rise
rise --help

For an external/team repo:

bash
cd ~/code/org/repo
rise adopt .
rise sync

For daily work:

bash
rise verify
rise mobile preflight      # if mobile repo
jj new main -m "feat: ..."

When To Use Rise

Use Rise when you want:

  • an opinionated operator layer over heterogeneous repos
  • clean PRs with private local tooling overlays
  • stronger mobile/testflight diagnostics
  • schema/sandbox/dev orchestration from one CLI surface

Use plain Flow-only setup when:

  • repo already has stable native workflows and minimal onboarding cost
  • you do not need overlay-based separation
  • your team explicitly avoids JJ overlay workflows

Relationship To Flow

Flow and Rise are complementary:

  • Flow is the general control plane (f tasks, f env, f commit, deploy/sync/invariants).
  • Rise is the repo/workflow acceleration layer for adoption, platform compile loops, mobile observability, and overlay workflows.

The practical entrypoint is still:

bash
f install rise

Then use rise where it adds leverage, while keeping Flow as your consistent command contract across projects.

References In Rise Docs

If you have access to the internal Rise repo docs, these are particularly relevant:

  • docs/adopt-guide.md
  • docs/workflow-guide.md
  • docs/rise-branch-workflow.md
  • docs/build-observability.md
  • docs/schema-guide.md
  • docs/sandbox-vibe.md
  • docs/rise-mobile-compat.md
  • docs/expo-identifiers.md