fleet-rlm 0.4.98

Recursive Language Models with DSPy + Modal and an integrated Web UI for secure long-context code execution

fleet-rlm

fleet-rlm is a Web UI-first recursive language model runtime for long-context code and document work. It ships a Modal-backed default runtime, an integrated FastAPI + WebSocket surface, packaged frontend assets, and an experimental Daytona workbench path that plugs into the same workspace instead of living as a separate product.

Docs | Contributing | Changelog

Why This Repo Exists

• Use a single workspace for long-context reasoning, chat turns, run inspection, and runtime diagnostics.
• Keep the default product path Modal-backed and chat-oriented.
• Expose an experimental Daytona pilot without forking the frontend or transport contract.
• Ship both a user-facing Web UI and integration surfaces for CLI, HTTP, WebSocket, and MCP workflows.

The supported app surfaces are RLM Workspace, Volumes, and Settings. Legacy taxonomy, skills, memory, and analytics routes now redirect to supported pages instead of remaining first-class product surfaces.

Quick Start

Add fleet-rlm to a uv-managed project and launch the Web UI:

# Create a project if you do not already have one
uv init

# Add fleet-rlm to the environment
uv add fleet-rlm

# Start the Web UI + API server
uv run fleet web

Open http://127.0.0.1:8000.

If you already have a uv project, skip uv init and just run uv add fleet-rlm.

Published installs already include built frontend assets, so end users do not need pnpm, vp, or a separate frontend build step.

Primary Workflows

Use the Web UI

uv run fleet web

This starts the main product surface with:

• RLM Workspace for chat and runtime execution
• Volumes for runtime-backed file browsing
• Settings for runtime configuration and diagnostics

Use terminal chat

uv run fleet-rlm chat --trace-mode compact

Run the API directly

uv run fleet-rlm serve-api --host 127.0.0.1 --port 8000

Enable MCP support

If you want the optional MCP server surface, install the extra first:

uv add "fleet-rlm[mcp]"
uv run fleet-rlm serve-mcp --transport stdio

Runtime Modes

fleet-rlm currently has two top-level runtime modes:

• modal_chat: the default product path
• daytona_pilot: the experimental workbench path

In the shared runtime contract:

• Modal requests can include execution_mode.
• Daytona requests can include repo_url, repo_ref, context_paths, and batch_concurrency.
• Daytona still uses the same websocket workspace and run-workbench flow, but it intentionally remains experimental.
• fleet-rlm daytona-rlm --max-depth remains available only as a deprecated compatibility flag for the CLI pilot path.

CLI Surfaces

This package exposes two command entrypoints:

• fleet: lightweight launcher for terminal chat and fleet web
• fleet-rlm: fuller Typer CLI for API, MCP, scaffold, and Daytona flows

Common commands:

# Web UI
uv run fleet web

# Terminal chat
uv run fleet
uv run fleet-rlm chat --trace-mode verbose

# FastAPI server
uv run fleet-rlm serve-api --port 8000

# Optional MCP server
uv run fleet-rlm serve-mcp --transport stdio

# Scaffold bundled Claude Code assets
uv run fleet-rlm init --list

# Experimental Daytona validation
uv run fleet-rlm daytona-smoke --repo https://github.com/qredence/fleet-rlm.git --ref main

# Experimental Daytona pilot
uv run fleet-rlm daytona-rlm \
  --repo https://github.com/qredence/fleet-rlm.git \
  --task "Summarize the tracing architecture" \
  --batch-concurrency 4

HTTP and WebSocket Contract

The current frontend/backend contract centers on:

• /health
• /ready
• GET /api/v1/auth/me
• GET /api/v1/sessions/state
• /api/v1/runtime/*
• POST /api/v1/traces/feedback
• /api/v1/ws/chat
• /api/v1/ws/execution

When AUTH_MODE=entra, HTTP and WebSocket access use real Entra bearer-token validation plus Neon-backed tenant admission. Runtime settings writes are intentionally limited to APP_ENV=local.

The canonical schema lives in openapi.yaml.

Source Development

From the repo root:

uv sync --all-extras --dev
uv run fleet web

Frontend contributors should use pnpm inside src/frontend:

cd src/frontend
pnpm install --frozen-lockfile
pnpm run dev
pnpm run check

This repo explicitly uses pnpm for frontend work even though the packaged frontend is built with Vite+ under the hood.

Validation

Repo-level validation:

make test-fast
make quality-gate
make release-check

Focused docs validation:

uv run python scripts/check_docs_quality.py
uv run python scripts/validate_release.py hygiene
uv run python scripts/validate_release.py metadata

Experimental Daytona Notes

Use this order for Daytona work:

1. Set DAYTONA_API_KEY, DAYTONA_API_URL, and optional DAYTONA_TARGET.
2. Run uv run fleet-rlm daytona-smoke --repo <url> [--ref <branch-or-sha>].
3. Only then run uv run fleet-rlm daytona-rlm [--repo <url>] [--context-path <path> ...] --task <text> ....

This repo treats DAYTONA_API_BASE_URL as a misconfiguration. Use DAYTONA_API_URL instead.

Documentation Map

• Documentation index
• Installation guide
• Developer setup
• CLI reference
• HTTP API reference
• Auth reference
• Frontend/backend integration
• Runtime settings
• Using the MCP server
• MLflow workflows