Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/shawnclybor/clybor-claude-toolingWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/shawnclybor/clybor-claude-tooling/ralph-loop)<a href="https://agentmods.dev/commands/shawnclybor/clybor-claude-tooling/ralph-loop"><img src="https://agentmods.dev/badge/commands/shawnclybor/clybor-claude-tooling/ralph-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/shawnclybor/clybor-claude-tooling/ralph-loop"><img src="https://agentmods.dev/badge/commands/shawnclybor/clybor-claude-tooling/ralph-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00028 | $0.00440 |
| Opus 5.5 | $0.00011 | $0.00176 |
| Sonnet 5.5 | $0.00006 | $0.00088 |
| Haiku 4.5 | $0.00003 | $0.00044 |
Grade A, and why
ralph-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
/ralph-loop
Start a Ralph loop. Implements the Ralph Wiggum technique: a single prompt re-fed to Claude on every session exit via a Stop hook, until Claude emits an exact completion-promise string or max-iterations is reached.
Usage
/ralph-loop "<prompt>" --completion-promise "<text>" --max-iterations <n>
Options:
--completion-promise <text>— exact string Claude emits to signal done (default:DONE)--max-iterations <n>— hard cap (default: 50; required for safety)
What happens
- Write
.ralph-loop/state.jsonwith the prompt, completion-promise, and iteration budget - Start working on the prompt
- When the session tries to exit,
.claude/hooks/ralph-stop.shfires - The hook checks the transcript for the completion-promise:
- Found → state cleared, exit allowed
- Not found AND under budget → state incremented, exit blocked, prompt re-fed
- Not found AND budget exhausted → state cleared, exit allowed with halt message
Cancel an active loop
Delete .ralph-loop/state.json to stop the loop on the next iteration:
rm .ralph-loop/state.json
When to use
Well-defined tasks with automatic verification (tests, linters, build), greenfield projects you can walk away from, anything where iterative refinement against deterministic checks is the right pattern.
When NOT to use
Tasks requiring human judgment, one-shot operations, unclear success criteria, production debugging (use /five-whys or the debugger agent instead).
Prompt writing — see the skill
The ralph-loop skill documents the prompt-writing best practices (clear completion criteria, incremental goals, self-correction, escape hatches). Read it before launching a long-running loop.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago Changed · +1 lines d346d9001e6f
- 1mo ago First seen · 48 lines · 28 tokens per session scan A 06129f0bc973
ralph-loop is a command published in the GitHub repository shawnclybor/clybor-claude-tooling (6 stars, last pushed 6d ago), licensed MIT. It adds 28 tokens to every session and 440 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
add-phase
Add phase to end of current milestone in roadmap.
insert-phase
Insert urgent work as decimal phase (e.g., 72.1) between existing phases.
remove-phase
Remove a future phase from roadmap and renumber subsequent phases.
research-phase
Research how to implement a phase (standalone — usually use /pbr:plan-phase instead).
reapply-patches
Reapply local modifications after a PBR update.
add-todo
Capture idea or task as todo from current conversation context.