메뉴
Meta-skill proactively used to clean, organize, and enforce structural standards within the .agent directory and antigravity-config. Triggers when the user asks to "organize", "clean", or "structure" the agent ecosystem.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Consolidated master domain for agentic arch.
Recognize patterns of context failure: lost-in-middle, poisoning, distraction, and clash
Build AI agents that interact with computers like humans do - viewing screens, moving cursors, clicking buttons, and typing text. Covers Anthropic's Computer Use, OpenAI's Operator/CUA, and open-source alternatives. Critical focus on sandboxing, security, and handling the unique challenges of vision-based control. Use when: computer use, desktop automation agent, screen control AI, vision-based agent, GUI automation.
Implement comprehensive evaluation strategies for LLM applications using automated metrics, human feedback, and benchmarking. Use when testing LLM performance, measuring AI application quality, or establishing evaluation frameworks.
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
Design short-term, long-term, and graph-based memory architectures
| name | agent-ops |
| description | Meta-skill proactively used to clean, organize, and enforce structural standards within the .agent directory and antigravity-config. Triggers when the user asks to "organize", "clean", or "structure" the agent ecosystem. |
| version | 1.0.0 |
| author | Antigravity |
[!IMPORTANT] This skill MUST be executed strictly under the Omni-Architect Agent Protocol v1.0. All tool executions, code modifications, and communications MUST adhere to the 13 core protocols.
agent-ops is the designated structural custodian of the Antigravity ecosystem. It operates primarily within the .agent directory (local) or antigravity-config (upstream template).
This skill teaches the agent how to audit the file tree, identify loose files, detect naming convention violations, and automatically reorganize them following the agent-workspace-standards.md policy.
Activate this sequence automatically when the user issues commands like:
enforce_structure()When activated, you must perform the following autonomous steps using your terminal/bash capabilities:
List all files in the root of the targeted .agent directory.
Identify any files that are NOT: README.md, INDEX.md, install.sh, .env, .gitignore.
Using standard mv commands, route the orphaned files to their corresponding domain based on the Global Rules:
Move Community/Standard Docs to /docs
mkdir -p docs
mv CODE_OF_CONDUCT.md docs/ 2>/dev/null
mv CONTRIBUTING.md docs/ 2>/dev/null
mv SECURITY.md docs/ 2>/dev/null
mv CHANGELOG.md docs/ 2>/dev/null
mv GEMINI.md docs/ 2>/dev/null
Move Loose Scripts to /scripts
mkdir -p scripts
mv *.py scripts/ 2>/dev/null
mv *.sh scripts/ 2>/dev/null
# Exclude install.sh intentionally
mv scripts/install.sh ./ 2>/dev/null
Move Orphaned Rules to /rules
mkdir -p rules
# If any ad-hoc rules were dumped in root
Review the subdirectories. If you spot CamelCase or spaces, alert the user and propose a mv rename to strict kebab-case.
Delete empty temporary files, failed script artifacts (test.py, output.txt), or .DS_Store cache fragments.
Once the mv sequences are complete, print a visual tree (ls -la or tree if available) to the user via the notify_user block, demonstrating the newly enforced state.