evalmonkey vs GSD (get-shit-done)

evalmonkey is an open-source framework for testing how LLM agents degrade under adversarial conditions. You run your agent against 10 standard datasets (GSM8K, ARC, HellaSwag, etc.) pulled automatically from HuggingFace, then apply chaos profiles that introduce realistic failure modes: malformed JSON schemas, artificial latency spikes, 429 rate-limit errors, context-window overflow, and prompt injection payloads. The key output is a degradation delta — evalmonkey shows you exactly how much your agent's accuracy drops under each failure type versus clean inputs. A model that scores 78% on GSM8K normally but drops to 31% when it gets a 429 mid-chain tells you something crucial about its error-recovery behavior that standard benchmarks completely miss. It supports OpenAI, Anthropic (via Bedrock and direct), Azure, GCP, and any Ollama-hosted model. Corbell-AI published this with a clear thesis: agents break in production for infrastructure reasons, not model reasons — and no existing benchmark tests that. evalmonkey was created today (April 17, 2026) and is still at 3 stars, but the core idea is genuinely novel in the evals space.

GSD (get-shit-done) is a meta-prompting and context engineering system for Claude Code that imposes software engineering discipline on AI-assisted development. It replaces ad-hoc prompting with a five-step methodology — initialize, discuss, plan, execute, verify — that keeps context fresh and quality high across long, complex projects. The system works by loading specialized documentation strategically: project vision, requirements, roadmaps, and research are injected at the right phases rather than dumped into a single bloated context window. Planning produces XML-formatted task trees with built-in verification steps, and execution happens in waves — parallel where dependencies allow, sequential where they don't. Quality gates automatically detect schema drift, security regressions, and scope creep before they compound into bigger problems. For teams that have experienced the quality degradation that hits around hour three of a long Claude Code session, GSD's architecture of fresh context windows per phase is the fix. A Quick Mode handles ad-hoc tasks without the full planning overhead, making it practical for both exploratory work and milestone-driven development. It's MIT-licensed, JavaScript-based, and designed for solo developers and small teams who want spec-driven development without enterprise process overhead.