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.
npx agentmods add commands/arc-mcp/arc-1/ralphex-plangit clone --depth 1 https://github.com/arc-mcp/arc-1Wrote 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/arc-mcp/arc-1/ralphex-plan)<a href="https://agentmods.dev/commands/arc-mcp/arc-1/ralphex-plan"><img src="https://agentmods.dev/badge/commands/arc-mcp/arc-1/ralphex-plan.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.08173 |
| Opus 5 | $0.00000 | $0.04086 |
| Sonnet 5 | $0.00000 | $0.01635 |
| Haiku 4.5 | $0.00000 | $0.00817 |
Grade A, and why
ralphex-plan 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.
How it starts
The opening of the file, as written. The whole thing — 407 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ralphex Plan Creator
Generate a structured implementation plan for the ralphex autonomous coding agent. The plan must be detailed enough for an autonomous agent (Claude Code) to execute each task without human guidance, in isolated sessions with fresh context.
Canonical plan-file format and ralphex behavior:
docs/ralphex.md. ralphex creates a git branch from the plan filename, runs each### Task N:in a fresh Claude Code session, runs## Validation Commandsafter each task, then runs a multi-agent review pipeline. The plan file is the only shared state between tasks.
The golden rule (why most plans fail): ARC-1 plans break when they assert SAP facts that aren't true on the target system — non-existent tool actions, wrong ADT endpoints/URLs, invented objectType/slash codes, XML response shapes ADT never emits, or behavior that differs by SAP release. A fresh-context agent has no way to know a fact is wrong; it implements the plan as written and ships the bug (a 200 OK does not mean a filter or type was honored). Every SAP-specific fact in the plan must be verified against the live system during research and cited as evidence (see Phase 1g). When in doubt, verify; never plan from memory of ADT internals.
Input
The user provides a feature/task description. This can be:
- A free-form description ("add caching for ADT responses")
- A reference to a gap analysis or roadmap item
- A GitHub issue
- A detailed requirements brief
If the description is vague, ask 1-2 targeted clarifying questions before proceeding. Don't over-interview — get enough to start researching.
Phase 1: Deep Research
Before writing a single line of the plan, research exhaustively. ralphex tasks execute in isolated Claude Code sessions with no shared context — every task must be self-contained. The quality of your research directly determines whether the autonomous agent succeeds or gets stuck.
1a. Read project guidelines
Read CLAUDE.md in full. Pay attention to:
- Codebase structure (the file tree)
- Key Files for Common Tasks table
- Code patterns (ADT client method, handler pattern, safety check)
- Testing conventions (vitest, mock patterns, fixture locations)
- Technology stack
- Configuration options (the config table — new features may need new flags)
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 First seen · 407 lines · 0 tokens per session scan A 78c9cd4bb387
ralphex-plan is a command published in the GitHub repository arc-mcp/arc-1 (177 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 8,173 tokens. 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.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
constitution
Create or update the project constitution from interactive or provided principle inputs.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.