Find agent skills by outcome

AGENTS.md — Universal Agent Instructions

> Works with: Claude Code, Cursor, Codex, Windsurf, and any AI coding agent.

C++: ROS2 C++ nodes

performance optimization

focus_areas: Edge AI for military robotics

context-aware vision systems

Edge deployment (TensorRT

ONNX

target_companies: [Anduril

Palantir

hardware: ESP32

RP2040

target_skills: Edge AI optimization

ROS2

Coding

programming_skills: Basic Python

C++

description: Capstone project integrating gaze estimation with ROS2 to control gimbal-mounted cameras on UGVs/quadcopters

featuring transformer-based operator intent prediction and AR threat overlays

Coding

languages: Python (intermediate)

Rust (basic)

12-Month AI and Computer Vision Roadmap for Defense Applications

{

- Key risks or dependencies to monitor (e.g.

bias checks)

- Instead

suggest non-AI or hybrid alternatives (e.g.

Data

- Suggested follow-up actions (prototype scope

data prep

- Recommendations for iteration (A/B testing

feedback loops)

### Version 1.5 (January 11

2026)

- Clear

grounded

- Lightweight/chat-only models (older 7B–13B models

mini variants) — usually lack depth/context/tool reliability

- Mistral Large 2 / Devstral — Very strong price/performance

fast

Data

- Llama 4 / Llama 405B variants (Meta) — Best open-source frontier performance

excellent for self-hosting

- Keep it minimal and adaptable

including placeholders for iteration or error handling

Creative

- Generate a simple

practical starter prompt

- Claude 4 Opus/Sonnet (Anthropic) — Exceptional long-form reasoning

writing quality

Productivity

- GPT-5 / o3 family (OpenAI) — Deepest reasoning on very complex structured tasks

best at following extremely long/complex instructions