Add MAINTAINING.md — a guide for editing the install prompt and setup docs#54
Merged
Conversation
… docs Consolidates the load-bearing invariants (previously split across CLAUDE.md and field-test/README.md) into one map a maintainer opens first: a traffic-light table of the five artifacts and their edit rules, the three-places-must-match rule for the prompt, the field-test gate to run before shipping, and the refusal-trigger don'ts. Adds discoverability pointers from CLAUDE.md (for the LLM) and a source-only HTML comment in GETTING-STARTED.md step 1 (for a colleague editing the onboarding prose) so both land on the guide before touching the prompt. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
Comprehensive guide clarifies editing rules for load-bearing docs. 🎯 Quality: 99% Elite · 📦 Size: Small 📈 This month: Your 218th PR — above team average · Averaging Excellent |
Contributor
Author
|
/review |
mariojgt
approved these changes
Jul 14, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
A colleague updating the onboarding steps needs to know — and their LLM needs to know — that the one-line install prompt is a load-bearing, adversarially-tuned artifact that must not be casually edited, while the surrounding prose is safe to improve. Today those rules exist but are split between
CLAUDE.md(terse, LLM-facing) andfield-test/README.md(the harness "why"). Neither is framed as "you're about to edit the docs — here's what's safe."The trap: "update the onboarding steps" sounds harmless, but step 1 of
GETTING-STARTED.mdcontains the prompt, which must stay byte-identical acrossREADME.md,GETTING-STARTED.md, andfield-test/prompt.txt.What
MAINTAINING.md(new) — the "guide for the guide": a traffic-light table of the five artifacts and their edit rules (🔴 don't touch / 🟠 edit-with-gate / 🟢 safe), the three-places-must-match rule, the field-test gate command to run before shipping, and the refusal-trigger don'ts. Wording is pulled from the existing invariants so nothing contradictsCLAUDE.mdorfield-test/README.md.CLAUDE.md— one-line pointer so the LLM lands on the guide.GETTING-STARTED.md— a source-only HTML comment in step 1 (invisible in the rendered onboarding) warning anyone editing the prompt block.Docs only — no code, no shipped-tarball content changed (
MAINTAINING.mdisn't inpackage.jsonfiles, so it stays dev-facing and out of the artifact agents audit).🤖 Generated with Claude Code