oh-my-claudecodevsSculptor
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
oh-my-claudecode
Yeachan Heo · Agent harness
OSSMCP
Panel
5.63 spec wins
- Reliability
- 5.2
- Usefulness
- 6.0
- Cost
- 5.8
- Longevity
- 5.3
“Has a Deep-Interview mode that asks you Socratic questions, making it the first agent to assign the human homework.”
Sculptor
Imbue · Agent harness
OSS
Panel
5.60 spec wins
- Reliability
- 5.2
- Usefulness
- 5.8
- Cost
- 6.8
- Longevity
- 4.7
“It ships a feature called CI Babysitter, which is the most accurate name anyone in this category has managed.”
Spec by spec
| Spec | oh-my-claudecode | Sculptor |
|---|---|---|
| Architecture | ||
| Category | Agent harness | Agent harness |
| Runs | local | local |
| Platforms | macos, linux, windows | macos, linux |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | Yes | No |
| MCP server | No | No |
| Capabilities | ||
| Runs terminal commands | Yes | Yes |
| Multi-file edits | Yes | Yes |
| Git operations | Yes | Yes |
| Browser control | No | No |
| Sandboxed execution | No | NoThe marketing page says every agent runs in its own container, but the help docs state the default workspace is a git worktree and describe a Docker container backend as experimental. |
| Multi-agent orchestration | Yes | Yes |
| Headless / CI mode | Yes | NoThere is no headless mode, but a CI Babysitter watches open pull requests and dispatches an agent to fix failing checks and merge conflicts. |
| Models | ||
| Backbone | Claude, Codex, Gemini, Antigravity, Grok, Cursor | Claude |
| Bring your own model | Yes | Yes |
| Local models | No | No |
| Cost | ||
| Pricing model | byok | byok |
| Starts at | $0/mo | n/a |
| Free tier | Yes | Yes |
| Bring your own key | Yes | Yes |
| Openness | ||
| Open source | Yes | Yes |
| License | MIT | MIT |
| GitHub stars | 39,522 | 234 |
Which one would each critic pick
| Critic | oh-my-claudecode | Sculptor | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 6.5 | 6.3 | oh-my-claudecode — Pick oh-my-claudecode if you run Claude Code all day and want plan, PRD, execute, verify and fix as a pipeline with Codex and Gemini as second opinions; pick Ruflo if you want swarms. |
| El Crítico | 5.3 | 5.3 | no preference |
| El Profesor | 5.0 | 6.5 | Sculptor — It drives the wrapped agent as a streaming-JSON process with the control protocol enabled and substitutes its own ask-user and plan tools, which is interception done properly. |
| La Inversora | 4.8 | 5.3 | Sculptor — This is a research lab's artefact rather than a company's product: no pricing page, a preview label, and a couple of hundred stars after a year. |
| La Jefa | 5.0 | 4.8 | oh-my-claudecode — tmux for team mode, the Linux flock utility for named autopilot profiles, and stop notifications wired to Telegram, Discord and Slack; sixty configurations, no admin; not yet. |
| El Hacker | 7.0 | 5.8 | oh-my-claudecode — MIT, .claude/omc.jsonc per project, ~/.config/claude-omc/config.jsonc per user, a .mcp.json, and OMC_STATE_DIR to keep state when a worktree dies; every knob is a file. |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.