Omg Executor — Development agent for Claude Code
Implements code changes for oh-my-grok tasks.
How to install Omg Executor
Installs to ~/.claude/agents/iml1s-oh-my-grok-omg-executor.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/ImL1s/oh-my-grok/HEAD/agents/omg-executor.md -o ~/.claude/agents/iml1s-oh-my-grok-omg-executor.md Restart Claude Code, or start a new session, for it to be picked up.
What Omg Executor does
name: omg-executor description: Implements code changes for oh-my-grok tasks. Use for write-heavy workstreams under ULW/RALPH. promptMode: extend permissionMode: default capabilityMode: read-write agentsMd: true disallowedTools:
- spawn_subagent
- run_terminal_command
- run_terminal_cmd tools:
- find_by_name
- grep_search
- view_file
- list_dir
- read_url_content
- search_web
- multi_replace_file_content
- replace_file_content
- write_to_file
- notebook_edit
Alternatives in Development
- Pua Action Executor — 普通执行 Agent:按任务说明完成代码/文档/配置改动,并输出候选结果;不做最终验收结论 19.5k ★
- Advisor — executor が返した advisor-request.v1 に対して方針だけ返す非実行 advisor 3.1k ★
- Workflow Executor — Implements single tasks atomically with verification and commit discipline 530 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Omg Orchestrator
Coordinates ULW/RALPH workstreams for oh-my-grok. Decomposes goals, spawns depth=1 Grok-native workers, integr
Omg Critic
Adversarial review of plans and code for oh-my-grok. Use under RALPLAN critic stage or ULW/RALPH review. Prefe
Ralph Coder
Implements code for a single Ralph user story. Writes production code and documentation only — does NOT write
Hephaestus
Autonomous deep implementation worker. Full tool access for complex coding tasks. Explores codebase patterns,
Coder
Implementation agent using TDD + Ralph Loop for feature development. Use proactively for implementation tasks
Forge Executor
Implements individual tasks from a frontier. Follows TDD when available, commits atomically, updates state. Di
Related Skills
Glm Delegate
Delegate token-heavy execution to swarms of GLM agents in disposable git worktrees — an Anthropic model writes
Claude Token Efficient
One CLAUDE.md file. Keeps Claude responses terse. Reduces output verbosity on heavy workflows. Drop-in, no cod
Mad Max
Unattended backlog-drain mode — runs executor → verifier → code-reviewer loops on all ready tasks, logs stuck/