TORCH

TORCH 0.1.0-alpha.3 · Experimental pilot

An AI team shaped by your project.

TORCH studies your repository and goals, then proposes the sessions your work needs. Choose Codex, Claude, Pi, or another local runtime for each role—and keep the team coordinated as the project changes.

Experimental alpha.3. The TORCH-on-TORCH self-host pilot is Codex-run. Built-in Codex, Claude, and Pi adapters exist; live mixed-provider swaps and production use remain unqualified.

Install TORCH Source and install guide

From project brief to checked work

  1. P
    Project + specificationTORCH maps code, tests, constraints, and stated goals.
  2. T
    Proposed teamYou review domains, responsibilities, and reporting lines.
  3. AI
    Your runtime for each sessionCodex, Claude, Pi, or a trusted local adapter.
    Direct peer links
  4. ↔
    Shared coordinationSessions exchange messages, status, tasks, and evidence.
  5. ✓
    Checked integrationWork converges through an authorized, serialized path.

Start with the repository. The proposed fleet stays yours to shape.

Built around your repo

Codex · Claude · Pi

Choose per session

Local-first control

Owner-shaped fleet

The project shapes the organization.

TORCH does not start from a fixed roster. It studies the repository and specifications you choose, then proposes a team you can split, merge, rename, or reject before anything is installed.

Understand

Read the real project

Architecture, tests, working practices, shared files, and your selected specifications become evidence for a proposed fleet.

Coordinate

Give each session a lane

Roles get explicit scope, exclusions, neighbors, checks, and worktree boundaries. Each session can use the runtime and model that fit its job.

Grow

Change shape when needed

As coordination grows, TORCH can recommend a lead or new specialist when there is real integration work to own. You review the evidence and approve every organization change.

Keep the right context close to the work.

Each long-lived session can focus on a coherent part of the project, while durable messages and shared records connect it to the rest of the fleet. The aim is less context thrash—not a set of agents trapped in silos.

Lower cache and context costs are a hypothesis TORCH still needs to measure.

Trust comes from evidence.

The self-host pilot uses the accepted immutable 0.1.0-alpha.3 artifact from source 28c. Later canonical source at 0c2 is not installed automatically. The fleet and its installed user-system schedule timer are manually paused; automatic AI wakes remain paused. Pipeline execution, isolated candidate-executor assurance, and production acceptance remain open.

CapabilityCurrent evidenceState
Project analysis and fleet designRepository/spec scenariosProven locally
Shared messages, status, and tasksDurable local control planeProven locally
Exact-commit checks and serialized integrationImplemented coordination foundationCandidate executor and fence assurance open
Codex, Claude, Pi, and local adaptersBuilt-in adapters; live mixed-provider swap unqualifiedLive provider gate open
TORCH-on-TORCH self-host pilotCodex-run; accepted alpha.3 artifact is source 28c; canonical source 0c2 is not auto-installedPilot paused; full qualification open
COMBATRIG importRead-only inventory; ownership review neededMigration gated

Install TORCH.

Experimental source install for developers. You need Git, Node.js 22 or later, and npm. Linux is the current self-host pilot environment; other platforms are not yet qualified.

Clone the rewrite branch, not the repository’s default branch. This installs the current development source, not the separately accepted pilot artifact. There is no stable release install promised here.

The isolated prefix below avoids overwriting an existing TORCH installation. The PATH change lasts for this terminal; add it to your shell configuration only if you want to keep using this source install.

git clone --branch rewrite/portable-agent-fleet \
  https://github.com/PR0M3TH3AN/TORCH.git
cd TORCH
npm ci
npm install --global --prefix "$HOME/.local/torch-source" .
export PATH="$HOME/.local/torch-source/bin:$PATH"
torch --help

Review source and dependencies before installing. These commands install the CLI only; they do not create or start a fleet. Already have TORCH? Use your existing installation’s reviewed update flow instead.

Then choose your project.

Point TORCH at a Git repository and, if useful, a product specification. It prepares an evidence brief for a fleet proposal; installing or starting the team waits for your review and approval.

torch bootstrap \
  --repo /path/to/project \
  --spec /path/to/product-spec.md \
  --output fleet-design-brief.json \
  --json

Review the team before it runs.

After installing the CLI, generate a baseline proposal from your project directory. Review domains, ownership, checks and unresolved gaps. In the proposal, set review.status to approved and record review.reviewedBy and review.reviewedAt only after your review.

The example selects Codex. Install and sign in to your chosen agent CLI separately; model access depends on your account. Each session’s runtime and model can be changed later.

Preview installation and worktree setup first. Setup creates configuration and worktrees but does not launch agents. Starting agents and installing host schedules are separate approvals.

Read the full setup, runtime configuration and uninstall guide.

cd /path/to/project
torch design --output fleet-proposal.json --json
# Review and approve fleet-proposal.json before continuing.
torch install --proposal fleet-proposal.json \
  --runtime codex --dry-run --json
torch install --proposal fleet-proposal.json \
  --runtime codex --yes --json
torch setup --dry-run --json
torch setup --yes --json
torch up --fresh --dry-run --json

The last command previews startup only. When you are ready to use your provider account, explicitly run torch up --fresh --yes. Check the plan and required gates first.