agentleFS
Sign inStart free
VepricovVepricov/claude-brainlab/rules/agents.md

claude-brainlab / rules

AGENTS.md35 starsChanged 2 months agoMITView on GitHub

Use a specialist when the work benefits from an independent review or a distinct area of expertise. A code change does not require a fixed sequence of planning, architecture, testing, and review agents. Handle small, clear edits directly. For delegated work, name the owned files or responsibility, expected result, and checks. Preserve other agents' edits and resources. Use the agents available in the current client; an unavailable agent name is not a reason to abandon work that can be completed…

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this agents.md into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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
  1. Distinctive:

    Use a specialist when the work benefits from an independent review or a distinct area of expertise.

    line 5
  2. Distinctive:

    All subagents run on Sonnet by default.

    line 14
  3. Distinctive:

    Run independent bounded tasks in parallel when delegation is allowed and useful work can continue locally.

    line 20

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
agents.md

Agent Orchestration

Choose help by the task

Use a specialist when the work benefits from an independent review or a distinct area of expertise. A code change does not require a fixed sequence of planning, architecture, testing, and review agents. Handle small, clear edits directly.

For delegated work, name the owned files or responsibility, expected result, and checks. Preserve other agents' edits and resources. Use the agents available in the current client; an unavailable agent name is not a reason to abandon work that can be completed locally.

Model for subagents (HARD RULE)

All subagents run on Sonnet by default. Владелец 02-11-2026: «все инджесты запускать на сонете, все агенты должны его по умолчанию использовать».

  • ~/.claude/settings.json sets CLAUDE_CODE_SUBAGENT_MODEL=sonnet, and every agent in ~/.claude/agents/ pins model: sonnet (except obsidian-hub-creator, which stays on Haiku).
  • Paper ingests (paper-ingest, want-2-read, any batch of papers) always run on Sonnet. When spawning one through the Agent tool, pass model: "sonnet" explicitly.
  • Override with another model only when the owner asks for it in that conversation.

Parallel execution

Run independent bounded tasks in parallel when delegation is allowed and useful work can continue locally. Keep dependent operations and shared-file edits sequential.

Available agents (~/.claude/agents/)

Research: literature-reviewer, paper-miner, kaggle-miner Dev: architect, build-error-resolver, code-reviewer, refactor-cleaner, tdd-guide, bug-analyzer, dev-planner Design: ui-sketcher, story-generator

Academic reviews and rebuttals use astar-paper-review and review-response directly. Their own workflows define any specialist roles they need.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free