LoopCode

LoopCode is an autonomous software engineering orchestrator built on top of OpenCode. It acts as the planning, state machine, knowledge, and verification layer that guides OpenCode through executing complex, multi-task software goals.

Disclaimer: This project is not affiliated with, sponsored by, or endorsed by the OpenCode/Anomaly team.

Key Features

1. Shared Memory Engine: All multi-agent communication is centralized through a unified SQLite persistence layer (SCHEMA_SQL embedded source of truth). Agents pass structured contracts asynchronously without direct coupling.
2. Semantic Knowledge & Code Graph: Fast local semantic embeddings using fastembed and sqlite-vec. Incremental codebase indexing parses classes, functions, and variables via tree-sitter, augmented by an LSP client.
3. Parallel Task Scheduling: Decomposes natural language goals into a Directed Acyclic Graph (DAG) of tasks. The GitWorktreeScheduler performs topological sorting based on file writeAllowlist permissions and runs independent batches concurrently in isolated Git worktrees.
4. 5-Layer Verification Engine: Validates task correctness via Compilation, Lint & Style, Unit Testing, Security Scanning (semgrep/trivy with regex fallbacks), and LLM Review Agent. Distinct skipped status reported when scripts are missing.
5. Failure Evidence Re-planning: If a task exhausts its execution retries, compilation errors, test failures, and reviewer notes are injected back into the Planner Agent to self-correct the task DAG.
6. Cost & Budget Limits: Enforces Monthly ➔ Goal ➔ Task spend caps. If breached, the system raises a notice (or opt-in gated rollback) and terminates with exit code 77.
7. Loop/Oscillation Prevention: Hashes state signatures (phase, task index, modified files, retry attempts) to detect infinite loops and prompts the user or event bus for resolution.
8. Catalog-Aware Model Router: Maps tasks to optimal models based on capability requirements, catalog availability, budget caps, and caching adjustments.

CLI Usage

Installation

1. Standalone Binary Installer (Recommended)

curl -fsSL https://raw.githubusercontent.com/arjun-vegeta/loopCode/main/install.sh | bash

2. Local Development Build

bun install
bun run build
bun run package   # Compiles standalone binary via bun build

Running a Goal

Decompose a natural language goal, plan, execute, and verify:

bun run src/index.ts "Add a new endpoint for user profile retrieval"

Command Options

• -d, --db <path>: Path to the SQLite database (default: loopcode.db).
• -r, --resume <sessionIdOrTaskId>: Session ID or Task ID to resume.
• --login: Start in-TUI provider authentication workflow.
• --headless: Run in non-interactive CI mode.

Exit Codes

Code	Meaning
0	Goal completed successfully
1	Fatal error
77	Budget limit exceeded
130	Execution interrupted

Configuration

LoopCode checks for overrides in ~/.loopcode/config.toml:

[model]
default      = "anthropic/claude-sonnet-4-6"
planning     = "anthropic/claude-opus-4-6"
execution    = "anthropic/claude-sonnet-4-6"
verification = "anthropic/claude-sonnet-4-6"

[budget]
maxMonthlyCostUsd = 100.0
maxGoalCostUsd    = 10.0
maxTaskCostUsd    = 2.0

[proxy]
enabled        = false
kind           = "antigravity"
port           = 8080
autoStart      = true
providerId     = "antigravity"

[ui]
theme = "auto"
ascii = false

[safety]
permissionMode           = "acceptEdits"
allowDestructiveRollback = false
maxParallelAgents        = 5

Running Verification & Tests

Run full verification suite (typecheck, lint, formatting, and unit tests):

bun run typecheck
bun run lint
bun run format:check
bun run test

License

This project is licensed under the MIT License - see the LICENSE file for details.