Skip to content

generic.md

Latest commit

 

History

History
46 lines (34 loc) · 2.71 KB

generic.md

File metadata and controls

46 lines (34 loc) · 2.71 KB

Install — Generic (Any Agent/Framework)

Recommended: Run ./install.sh from the repo root for automated installation. Supports Claude Code, Gemini CLI, Codex CLI, Cursor, OpenCode, and OpenClaw. The manual steps below are for reference or troubleshooting.

Universal principle

Attention Allocation is a cognitive base — it changes how the agent allocates cognitive resources, not what tools it uses. Installation means injecting cognitive-protocol.md into the agent's always-on instructions (system prompt, rules file, or configuration).

Platform mapping

Platform Where to inject File to use
Claude Code ~/.claude/attention-allocation.md + ref in CLAUDE.md cognitive-protocol.md
Codex Prepend to AGENTS.md cognitive-protocol.md
Gemini system_instruction field cognitive-protocol.md
Cursor Prepend to .cursorrules cognitive-protocol.md
ChatGPT Custom Instructions -> system prompt cognitive-protocol.md
LangChain System message in chain cognitive-protocol.md
AutoGPT / CrewAI Agent system prompt cognitive-protocol.md
Any other Find the "always-on instructions" config and inject there cognitive-protocol.md

Step by step

  1. Locate your agent's system prompt or always-on rules file
  2. Copy the contents of cognitive-protocol.md (~30 lines)
  3. Paste it into the system prompt, BEFORE any domain-specific instructions
  4. If using First Principles, place it first, then Attention Allocation

File selection guide

Need File Size
Minimal install (core rules only) cognitive-protocol.md ~30 lines
Full reference framework + SKILL.md ~160 lines
Anti-pattern detection + anti-patterns.md ~130 lines
Teaching examples + examples.md ~130 lines

For most agents, cognitive-protocol.md alone is sufficient. The additional files are reference material for when the agent needs deeper guidance.

Troubleshooting

  • Agent ignores the rules: Move cognitive-protocol.md content to the beginning of the system prompt, not the end. Most models weight earlier instructions more heavily.
  • Agent still spreads attention evenly: This is the hardest habit to break. Add the anti-pattern checklist from SKILL.md section 3 to reinforce the filtering behavior.
  • Rules conflict with domain instructions: Attention Allocation should never conflict — it changes resource allocation, not output format. If conflict appears, the domain instruction likely assumes even distribution is desirable. That assumption is worth questioning.
  • Context window pressure: cognitive-protocol.md is ~30 lines. If that's too much, something else in your prompt needs trimming first.