brain
Skill1 starsChanged 4 months agoMITView on GitHub
Boot into the CLAUDE-BRAIN inheritance vault (E:/PRAGYESH - WORK/CLAUDE-BRAIN) before doing any work. Use at the start of any session or task in E:/PRAGYESH - WORK — building, fixing, auditing, researching, designing. Loads the vault's core behavior contract, routes to the ONE workflow doc matching the task, and enforces the sleep protocol (write-back) before ending.
25 other projects use a version of this.
Adopted by
Most recently taken up by
- tyler-berggren/tb-claude-kittheir own version, first seen Nov 12, 2026
- jzhao234/big-braintheir own version, first seen Nov 8, 2026
- jonmagic/skillstheir own version, first seen Nov 8, 2026
- keli-wen/agy-stafftheir own version, first seen Nov 8, 2026
- Aiqbee/aiqbee-aitheir own version, first seen Nov 8, 2026
- SeanningTatum/brain-axitheir own version, first seen Nov 1, 2026
Other versions25 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 25
- keli-wen/agy-staff★ 873same name
- H-mmer/pentest-agents★ 815same name
- Anantys-oss/koan★ 49same name
- sophonfinance-wq/finance-automation-portfolio★ 16same name
- henriquecaner/brain-agent★ 13same name
- Aiqbee/aiqbee-ai★ 7same name
- jonmagic/skills★ 6same name
- tonydzi/second-brain-starter-kit★ 6same name
- aidemd-mcp/server★ 4same name
- gufranco/claude-engineering-rules★ 3same name
15 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- skills/nirmaan/CLAUDE.mdCLAUDE.md
- fable-brainSkill
- full-pictureSkill
- graphifySkill
- kimi-webbridgeSkill
- neuronSkill
- nirmaanSkill
- no-slopSkill
- researchSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 2 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 20 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 21
First tenure only: if you (this model identity) have never booted the vault before, also read craft\C00-READ-ME-FIRST.md and …
- Very specific: line 25
If the task concerns an existing project: read its card in _REGISTRY\ (find it via _REGISTRY\INDEX.md), plus _REGISTRY\TIP-ECOSYSTEM.md if …
- Very specific: line 28
Route the task: read workflows\INDEX.md and load the ONE W-doc whose trigger matches the task.
- Very specific: line 31
No W-doc matches → read craft\C13-HOW-I-THINK.md and run its meta-procedure: do the task with the ten patterns, then WRITE the missing …
- Distinctive: line 47
Update in place; never fork or copy the vault.
- Distinctive: line 50
If a vault instruction conflicts with reality (file moved, API changed), fix the vault as part of the task and note the fix to the user.
“Common” is measured: kept by at least 6 in 10 of the 20 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/brain — Boot the CLAUDE-BRAIN vault
You are inheriting a knowledge vault written by a predecessor model for this specific user (a non-technical founder). The vault compensates for whatever you lack — follow its procedures over your own instincts wherever they conflict.
Vault root: E:\PRAGYESH - WORK\CLAUDE-BRAIN
Boot sequence (do exactly this, in this order)
- Read
HOME.mdat the vault root. It is the entry point and defines the four layers. Obey its load order. - Read
10-OPERATING-PARAMETERS.mdand set the dials for the task at hand. State your dial settings in one line to the user. - First tenure only: if you (this model identity) have never booted the
vault before, also read
craft\C00-READ-ME-FIRST.mdandcraft\C11-TIER-GAP-COMPENSATION.mdnow, and skim the craft titles. Craft docs are read once per model tenure, not per session. - If the task concerns an existing project: read its card in
_REGISTRY\(find it via_REGISTRY\INDEX.md), plus_REGISTRY\TIP-ECOSYSTEM.mdif the project is part of TIP. Follow the linked notes per the W83 wake-up protocol. - Route the task: read
workflows\INDEX.mdand load the ONE W-doc whose trigger matches the task. Do NOT load multiple W-docs or browse the vault — the index exists so you load ~3–5 files total, never all of them.- No W-doc matches → read
craft\C13-HOW-I-THINK.mdand run its meta-procedure: do the task with the ten patterns, then WRITE the missing W-doc and add its INDEX line.
- No W-doc matches → read
- Do the work under the loaded procedure. Core rules that always apply:
- Final gate (
08-FINAL-GATE.md) runs on every answer before sending. - Forbidden zone (payments, auth, security rules, migrations, deletes, cross-repo writes): propose, never execute without explicit approval.
- Delegating to a cheap model → build the envelope from
workflows\W84, orchestrate perworkflows\W94.
- Final gate (
- Sleep protocol (mandatory, before ending the session): per W83, write
back decisions, learnings, and state changes to the matching registry card
or W-doc. Any mistake the user corrected becomes a rule in the matching file
(template H in
09-PROMPT-TEMPLATES.md) — update in place, never duplicate. A session that doesn't write back steals from the next one.
Rules about the vault itself
- Update in place; never fork or copy the vault.
- New links use
[[wikilinks]]matching exact filenames (no extension), added under the## Linkedfooter of each touched file. - If a vault instruction conflicts with reality (file moved, API changed), fix the vault as part of the task and note the fix to the user.
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.

