{"id":46968338,"url":"https://github.com/computerlovetech/ralphify","last_synced_at":"2026-04-01T17:33:22.618Z","repository":{"id":343245310,"uuid":"1176039238","full_name":"computerlovetech/ralphify","owner":"computerlovetech","description":"Stop stressing over not having an agent running. Ralph is always running 🏃‍♂","archived":false,"fork":false,"pushed_at":"2026-03-31T19:06:00.000Z","size":2999,"stargazers_count":51,"open_issues_count":8,"forks_count":3,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-03-31T20:39:29.496Z","etag":null,"topics":["ralph","ralph-loop","ralph-wiggum"],"latest_commit_sha":null,"homepage":"https://ralphify.co/docs/","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/computerlovetech.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"docs/contributing/codebase-map.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"docs/agents.md","dco":null,"cla":null}},"created_at":"2026-03-08T14:35:13.000Z","updated_at":"2026-03-31T19:06:03.000Z","dependencies_parsed_at":"2026-03-13T13:00:45.795Z","dependency_job_id":null,"html_url":"https://github.com/computerlovetech/ralphify","commit_stats":null,"previous_names":["computerlovetech/ralphify"],"tags_count":16,"template":false,"template_full_name":null,"purl":"pkg:github/computerlovetech/ralphify","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/computerlovetech%2Fralphify","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/computerlovetech%2Fralphify/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/computerlovetech%2Fralphify/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/computerlovetech%2Fralphify/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/computerlovetech","download_url":"https://codeload.github.com/computerlovetech/ralphify/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/computerlovetech%2Fralphify/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31290537,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-01T13:12:26.723Z","status":"ssl_error","status_checked_at":"2026-04-01T13:12:25.102Z","response_time":53,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ralph","ralph-loop","ralph-wiggum"],"created_at":"2026-03-11T11:21:56.427Z","updated_at":"2026-04-01T17:33:22.610Z","avatar_url":"https://github.com/computerlovetech.png","language":"Python","readme":"\u003cp align=\"center\"\u003e\n  \u003cimg src=\"cli-graphic.png\" alt=\"ralphify\" /\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://pypi.org/project/ralphify/\"\u003e\u003cimg src=\"https://img.shields.io/pypi/v/ralphify?color=blue\" alt=\"PyPI version\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://pypi.org/project/ralphify/\"\u003e\u003cimg src=\"https://img.shields.io/pypi/pyversions/ralphify\" alt=\"Python versions\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://github.com/computerlovetech/ralphify/blob/main/LICENSE\"\u003e\u003cimg src=\"https://img.shields.io/github/license/computerlovetech/ralphify\" alt=\"License\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://ralphify.co/docs/\"\u003e\u003cimg src=\"https://img.shields.io/badge/docs-ralphify.co%2Fdocs-blue\" alt=\"Documentation\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\nPut your AI coding agent in a `while True` loop and let it ship.\n\nRalphify is a minimal harness for running autonomous AI coding loops, inspired by the [Ralph Wiggum technique](https://ghuntley.com/ralph/). The idea is simple: pipe a prompt to an AI coding agent, let it do one thing, commit, and repeat. Forever. Until you hit Ctrl+C.\n\n```\nwhile :; do cat RALPH.md | claude -p ; done\n```\n\nRalphify wraps this pattern into a proper tool with commands, iteration tracking, and clean shutdown.\n\n## Install\n\n```bash\nuv tool install ralphify    # recommended\n```\n\nOr if you don't have `uv`:\n\n```bash\npipx install ralphify       # isolated install via pipx\npip install ralphify         # plain pip (use a virtualenv or --user)\n```\n\nAny of these gives you the `ralph` command.\n\n## Quickstart\n\nA ralph is a directory with a `RALPH.md` file. Scaffold one:\n\n```bash\nralph init my-ralph\n```\n\nThen edit `my-ralph/RALPH.md`:\n\n```markdown\n---\nagent: claude -p --dangerously-skip-permissions\ncommands:\n  - name: tests\n    run: uv run pytest\n---\n\nYou are an autonomous coding agent working in a loop.\n\n## Test results\n\n{{ commands.tests }}\n\nIf any tests are failing, fix them before continuing.\n\n## Task\n\nImplement the next feature from the TODO list.\n```\n\nRun it:\n\n```bash\nralph run my-ralph           # Starts the loop (Ctrl+C to stop)\nralph run my-ralph -n 5      # Run 5 iterations then stop\n```\n\n### What `ralph run` does\n\nEach iteration:\n1. **Runs commands** — executes all commands, captures output\n2. **Assembles prompt** — reads RALPH.md body, replaces `{{ commands.\u003cname\u003e }}` placeholders with output\n3. **Pipes to agent** — executes the agent command with the assembled prompt on stdin\n4. **Repeats** — goes back to step 1\n\n### What it looks like\n\n```\n$ ralph run my-ralph -n 3\n\n▶ Running: my-ralph\n  1 command · max 3 iterations\n\n── Iteration 1 ──\n  Commands: 1 ran\n✓ Iteration 1 completed (52.3s)\n\n── Iteration 2 ──\n  Commands: 1 ran\n✗ Iteration 2 failed with exit code 1 (23.1s)\n\n── Iteration 3 ──\n  Commands: 1 ran\n✓ Iteration 3 completed (41.7s)\n\n──────────────────────\nDone: 3 iterations — 2 succeeded, 1 failed\n```\n\n## The technique\n\nThe Ralph Wiggum technique works because:\n\n- **One thing per loop.** The agent picks the most important task, implements it, tests it, and commits. Then the next iteration starts fresh.\n- **Fresh context every time.** No context window bloat. Each loop starts clean and reads the current state of the codebase.\n- **Progress lives in git.** Code, commits, and a plan file are the only state that persists between iterations. If something goes wrong, `git reset --hard` and run more loops.\n- **The prompt is a tuning knob.** When the agent does something dumb, you add a sign. Like telling Ralph not to jump off the slide — you add \"SLIDE DOWN, DON'T JUMP\" to the prompt.\n\nRead the full writeup: [Ralph Wiggum as a \"software engineer\"](https://ghuntley.com/ralph/)\n\n## Core concepts\n\nA **ralph** is a directory containing a `RALPH.md` file. That's it. Everything the ralph needs lives in that directory.\n\n```\nmy-ralph/\n├── RALPH.md              # the prompt (required)\n├── check-coverage.sh     # script (optional)\n├── style-guide.md        # reference doc (optional)\n└── test-data.json        # any supporting file (optional)\n```\n\n**RALPH.md** is the only file the framework reads. It has YAML frontmatter for configuration and a body that becomes the prompt:\n\n| Frontmatter field | Required | Description |\n|---|---|---|\n| `agent` | Yes | The agent command to run |\n| `commands` | No | List of commands (name + run) whose output fills `{{ commands.\u003cname\u003e }}` placeholders |\n| `args` | No | Declared argument names for `{{ args.\u003cname\u003e }}` placeholders |\n| `credit` | No | Append co-author trailer instruction to prompt (default: `true`) |\n\n**Commands** run before each iteration. Their output replaces `{{ commands.\u003cname\u003e }}` placeholders in the prompt. Use them for test results, git history, lint output — anything that changes between iterations.\n\n**No project-level configuration.** No `ralph.toml`. No config files. A ralph is fully self-contained.\n\n## AI-guided setup\n\n```bash\nralph new my-task\n```\n\nLaunches an interactive agent conversation to scaffold a new ralph with the right commands and prompt for your project.\n\n## Install ralphs from GitHub\n\n```bash\nralph add owner/repo              # Install all ralphs from a repo\nralph add owner/repo/my-ralph     # Install a specific ralph by name\n```\n\nInstalls ralphs to `.ralphify/ralphs/` so you can run them by name with `ralph run`.\n\n## Documentation\n\nFull documentation at **[ralphify.co/docs](https://ralphify.co/docs/)** — getting started tutorial, prompt writing guide, cookbook, and troubleshooting.\n\n## Requirements\n\n- Python 3.11+\n- [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) (or any agent CLI that accepts piped input)\n\n## License\n\nMIT\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcomputerlovetech%2Fralphify","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcomputerlovetech%2Fralphify","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcomputerlovetech%2Fralphify/lists"}