lazyFrogLOLCommunity·35 Agent Skills Included

Harness_Engineering

Autonomous long-running coding, testing, and terminal task execution

Runs long-horizon development tasks end to end: planning, contract negotiation, coding, browser testing, scoring, and iterative refinement. Eliminates premature task abandonment through context reset, anxiety detection, and loop-detection middleware. Includes 35 ready-made task skills covering terminal benchmarks, code repair, security analysis, and system configuration.

127|27

↗

npx skills add lazyFrogLOL/Harness_Engineering --all -g -y

All Skills in This Repository (35)

Pure Emerald Level Indicators

📦 In Repo

lazyFrogLOL

path-tracing

Analyze target images to extract rendering parameters and reproduce ray-traced scenes.

video-processing

Edit, convert, and extract media from video files using FFmpeg commands.

install-windows-3.11

Install Windows 3.11 in QEMU with VNC and noVNC remote access.

chess-best-move

Detect chess pieces from board images and compute best moves with Stockfish.

reshard-c4-data

Reshard large datasets into size-limited and count-limited shards with metadata.

gcode-to-text

Decode text embedded in G-code toolpaths by analyzing X/Y coordinate geometry.

vulnerable-secret

Extract hidden secrets from vulnerable executables using reverse engineering and exploitation techniques.

compile-compcert

Build CompCert from source using opam-managed Coq, OCaml, and Menhir dependencies.

train-fasttext

Guide FastText supervised text classification training under accuracy and size constraints.

mailman

Configure Postfix and Mailman3 mailing lists with LMTP delivery verification.

schemelike-metacircular-eval

Implement a metacircular Scheme evaluator with closures, environments, and special forms.

winning-avg-corewars

Create CoreWars warriors meeting win-rate targets against specified opponents.

Community

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install Harness_Engineering?▼

Run `npx skills add lazyFrogLOL/Harness_Engineering --all -g -y` in your terminal to install all skills in this suite globally.

What does Harness_Engineering do?▼

It turns a single sentence request into a fully autonomous workflow where a planner, builder, and evaluator plan, code, test in a real browser, and iterate until quality targets are met.

Which models and providers does it support?▼

It is built in pure Python on the OpenAI-compatible API, so it works with any model provider without requiring a specific vendor SDK.

Can it run Terminal-Bench 2.0 benchmarks?▼

Yes. It ships with a Harbor adapter, 35 task-specific skills, and a benchmark script, and it scored 43.8% on the Terminal-Bench 2.0 leaderboard.

Do I need coding experience to use it?▼

Basic Python setup is required to run the harness itself, but once configured it completes tasks autonomously from plain-English instructions.