run

run is a command for coding agents from Rubiss-Projects/spec-kit-ralph. It costs 7 tokens per session (1,631 once invoked), scanned A, original, MIT.

A command that starts the Ralph autonomous implementation loop, which repeatedly selects work from a tasks file and runs an implementation process. It accepts options such as iteration limits, model, agent command-line tool, and verbose output.

In plain words
What is it for?
Use it to launch Ralph with chosen settings and let the orchestrator select the next task from tasks.md.
Why use it?
It handles prerequisite checks and configuration in one launcher, while warning that free-form task text is ignored.

Command

Install

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.

agentmods
npx agentmods add commands/rubiss-projects/spec-kit-ralph/run
Clone the repo
git clone --depth 1 https://github.com/Rubiss-Projects/spec-kit-ralph

Wrote 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.

agentmods badge for run

README.md
[![agentmods](https://agentmods.dev/badge/commands/rubiss-projects/spec-kit-ralph/run.svg)](https://agentmods.dev/commands/rubiss-projects/spec-kit-ralph/run)
Your own site
<a href="https://agentmods.dev/commands/rubiss-projects/spec-kit-ralph/run"><img src="https://agentmods.dev/badge/commands/rubiss-projects/spec-kit-ralph/run.svg" alt="Measured on agentmods" height="20"></a>
Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,631 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00007 $0.01631
Opus 5 $0.00003 $0.00816
Sonnet 5 $0.00001 $0.00326
Haiku 4.5 $0.00001 $0.00163

Measured 3d ago against content hash bb44cc2c7c71, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

run 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 3d 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.

commands/run.md · 117 lines

How it starts

The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.

User Input

$ARGUMENTS

You MUST treat the user input as launcher arguments only.

Recognized launcher arguments are:

  • --max-iterations N or -n N
  • --model MODEL or -m MODEL
  • --agent-cli CLI
  • --verbose or -v

Free-form requests such as "Implement US1", "do the next story", or "fix the tasks" are not instructions for this command to execute work directly. If free-form text is present:

  1. Print a short warning that the text is ignored by speckit.ralph.run because Ralph selects work from tasks.md.
  2. Continue launching the Ralph orchestrator with the resolved configuration.

This command MUST NOT implement tasks, edit project files, mark checkboxes, create commits, or run /speckit.ralph.iterate inline. Its only job is prerequisite validation, configuration resolution, and launching the orchestrator script.

Purpose

This command is a thin launcher for the ralph loop orchestrator. It validates prerequisites, resolves configuration, and launches the platform-appropriate orchestrator script in a visible terminal for the user to monitor. Once the script is launched, this command's job is done — it exits immediately and does not wait for the script to complete.

Outline

  1. Parse launcher arguments only from $ARGUMENTS:
    • --max-iterations N or -n N (default: from config or 10)
    • --model MODEL or -m MODEL (default: from config or claude-sonnet-4.6)
    • --agent-cli CLI (default: from config or copilot; supported: copilot, codex, claude)
      • For copilot, resolve the registered Spec Kit command/skill name from .specify/integration.json. Dot separator uses --agent speckit.ralph.iterate; dash/skills mode invokes /speckit-ralph-iterate in the prompt. Spec Kit integration options such as --skills are not passed as Copilot runtime flags.
    • --verbose or -v (default: false)
    • Ignore non-flag free-form text after printing the warning described above
    • Stop with a clear error for unknown flags or malformed flag values

Read the full file on GitHub · 117 lines

Changes

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.

  1. 3d ago First seen · 117 lines · 7 tokens per session scan A bb44cc2c7c71

Subscribe to this mod's changes

run is a command published in the GitHub repository Rubiss-Projects/spec-kit-ralph (24 stars, last pushed 10d ago), licensed MIT. It adds 7 tokens to every session and 1,631 once invoked, about $0.0000 per session on Opus 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 2026-08-30.