A documentation quality methodology that spawns fresh agents to validate whether docs are actually usable.
The Problem
Documentation written by the person who built the thing is almost always incomplete. They fill in gaps unconsciously. They assume context. They skip "obvious" steps.
RTFM Testing fixes this by spawning a fresh agent with zero context and asking: can you complete this task using only the docs?
When to Use
Before publishing docs, READMEs, tutorials, or setup guides
When users report confusion but you can't see why
After major refactors to validate docs still work
As part of CI for documentation-heavy projects
How It Works
Identify the task — What should someone be able to do after reading the docs?
Bundle the docs — Collect all relevant documentation (and nothing else)
Spawn a fresh tester — Use the TESTER.md prompt with sessions_spawn
Analyze failures — Every confusion point is a doc bug
Fix and repeat — Update docs, respawn, retest until clean
Usage
sessions_spawn(
task: "Complete the following task using ONLY the provided documentation. [TASK DESCRIPTION]\n\n---\n\n[PASTE DOCS HERE]",
agentId: "default",
label: "rtfm-test"
)
Or use the full TESTER.md prompt for more structured output.
Metrics
Cold Start Score — Number of spawn cycles until task completion (lower = better docs)
Docs only — No external knowledge, no "common sense"
Failures are signal — Every stumble is actionable feedback
Files
SKILL.md — This file
TESTER.md — System prompt for the fresh agent
GAPS.md — Output format specification
1---2name: rtfm-testing3description: RTFM Testing4---5# RTFM Testing67A documentation quality methodology that spawns fresh agents to validate whether docs are actually usable.89## The Problem1011Documentation written by the person who built the thing is almost always incomplete. They fill in gaps unconsciously. They assume context. They skip "obvious" steps.1213RTFM Testing fixes this by spawning a fresh agent with zero context and asking: can you complete this task using only the docs?1415## When to Use1617- Before publishing docs, READMEs, tutorials, or setup guides18- When users report confusion but you can't see why19- After major refactors to validate docs still work20- As part of CI for documentation-heavy projects2122## How It Works23241. **Identify the task** — What should someone be able to do after reading the docs?252. **Bundle the docs** — Collect all relevant documentation (and nothing else)263. **Spawn a fresh tester** — Use the TESTER.md prompt with `sessions_spawn`274. **Analyze failures** — Every confusion point is a doc bug285. **Fix and repeat** — Update docs, respawn, retest until clean2930## Usage3132```33sessions_spawn(34 task: "Complete the following task using ONLY the provided documentation. [TASK DESCRIPTION]\n\n---\n\n[PASTE DOCS HERE]",35 agentId: "default",36 label: "rtfm-test"37)38```3940Or use the full TESTER.md prompt for more structured output.4142## Metrics4344- **Cold Start Score** — Number of spawn cycles until task completion (lower = better docs)45- **Gap Count** — Number of `[GAP]` reports per run46- **Gap Categories** — Missing steps, unclear language, wrong assumptions, missing prerequisites4748## Key Principles49501. **No hints** — Don't help the tester. Let it fail.512. **Literal reading** — Tester must not infer or guess523. **Docs only** — No external knowledge, no "common sense"534. **Failures are signal** — Every stumble is actionable feedback5455## Files5657- `SKILL.md` — This file58- `TESTER.md` — System prompt for the fresh agent59- `GAPS.md` — Output format specification
Run npx skillmds@latest add johnalbertini14-glitch/rtfm-testing in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
RTFM Testing It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: CAUTION, Skill Scanner: PASS. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
johnalbertini14-glitch (@johnalbertini14-glitch) published this skill. Their other Agent Skills are listed on their SkillMD profile.
johnalbertini14-glitch