clean-room-documentation
Skill16 starsChanged 6 months agoMITView on GitHub
Write safe architecture documentation without copying source.
In this repository8 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- architecture-synthesisSkill
- batch-planningSkill
- continue-batch-cyclesSkill
- index-fresh-projectSkill
- mirrored-file-docsSkill
- repair-consistencySkill
- verificationSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 66
Use medium detail for adapters, handlers, providers, renderers, transport layers, API wrappers, integrations, support modules, CLI helpers, …
- Distinctive: line 68
Use compact detail only for clearly simple constants, wrappers, aliases, formatting helpers, passthrough exports, one-function utilities, …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
clean-room-documentation
Clean-Room Rules
Absolutely never copy source code, implementation line-by-line, algorithms line-by-line, raw code into rebuild docs, or create rewritten-source files.
Allowed: names, paths, public API names, architecture descriptions, responsibilities, dependencies, relationships, data flow, execution flow, state transitions, permission boundaries, rebuild concepts, and behavioral descriptions.
This is architecture extraction, not cloning.
File Documentation Template
<relative/source/path>
Project: project_atlas
Batch: BATCH_####
Status: Draft | Needs Verification | Verified | Needs Rework
Purpose
Main Responsibilities
Key Exports
Inputs / Outputs
Imports / Dependencies
Referenced Modules
Relationships
Architectural Role
Clean-Room Rebuild Notes
Unknowns / Follow-up
Optional sections when helpful:
Important Types / Interfaces
Important Functions / Methods
State Mutations
Permission / Safety Behavior
Tool Calls / External Effects
Execution Flow
Error Handling
Test Coverage / Fixtures
Comparison Notes
Documentation Depth Rules
Documentation depth is based on architectural importance, not file size.
Deeply document anything affecting orchestration, execution, permissions, prompts, memory, tools, state, model routing, agent behavior, control flow, lifecycle, background processing, command routing, file operations, subprocesses, bridge/IPC, MCP, user safety, or rebuild-critical behavior.
Use medium detail for adapters, handlers, providers, renderers, transport layers, API wrappers, integrations, support modules, CLI helpers, and behavior-revealing test utilities.
Use compact detail only for clearly simple constants, wrappers, aliases, formatting helpers, passthrough exports, one-function utilities, and static metadata with no behavioral effect.
A single file may contain deep, medium, and compact sections. If uncertain, document more, not less.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.