Agentic development lint-fix-loop kit: session-scoped file tracking and a capped Stop-hook fix loop for Claude Code plugins.

Staged workflow for AI agents — code, writing, research, operations: plan, human approval, build, independent review, bounded fix loop.

Built-in BYOK LLM fix loop for align architecture-conformance violations (align agent run).

Marigold: the engine that writes test proposals. Tierless kill-matrix fix loop, light check and exit proof. "name" will publish under the @abloh scope.

The DoubleOh fix loop as an MCP server: skills_for, request_fix, report_skill.

GateTest — 121-module QA gate with MCP. Eyes (screenshot live pages), ears (production errors), hands (verify fixes worked). Security, supply chain, AI safety, mutation testing. Iterative Claude fix loop. Replaces SonarQube + Snyk + ESLint + 10 others.

CI doctor plugin for DeepSeek Harness: watch GitHub Actions failures, diagnose logs into structured findings, and drive the fix loop

DSH multi-model orchestration plugin: plan / execute / review pipeline with auto fix-loop, budget and degradation

Real-time LSP diagnostics for DeepSeek Harness — model-visible lsp_diagnostics / lsp_workspace_errors / lsp_hover tools plus an auto-injected post-edit diagnostics section, closing the edit → diagnostics → fix loop.

AI CLI Agent with built-in 9Router — REPL, multi-LLM, auto-fix loop, AI Router + Dashboard

Scaffold a GitHub Actions AI test-fix loop workflow into your repo

LLM-assisted validation → fix loop with deterministic verification

Single-binary coding agent: zero-setup codebase index, cost-aware model routing, worktree agents, auto-fix loop, voice cloning, autonomous browser, ACP for editors. Claude, Ollama, OpenAI-compatible.

Automated SonarQube PR fix loop for Claude Code, Gemini CLI, and Codex CLI — iterative fix, commit, and re-check until passing

Self-validating AI code agent CLI: per-worktree services, agent-driven validate-fix loop, and PR review mode.

A shim for the setImmediate efficient script yielding API

Kalmia eval optimizer — threshold-driven prompt fix loop over your own eval suite

Lock on asynchronous code

Abloh sandbox: runs a repository's suite inside an immutable, integrity-locked container and reports what happened. The container core the fix loop proves generated tests with.

Utility to fix TypeScript declarations when using default exports in CommonJS.