Brief: Initializes the Ourobor OS wiki structure in a target project and wires the LLM maintenance protocol into instruction files.

Overview

ouro/scripts/bootstrap.py is a one-time setup script run from a target project's root. It creates the ouro/wiki/ directory tree, copies the template skeleton from the installed skill, and appends the maintenance protocol to any LLM instruction files it finds.

Functions

detect_llm_environment()

Inspects the current working directory for known LLM config directories and returns a list of detected environments plus the primary instruction filename.

@snippet detect-llm-env

if (cwd / '.claude').exists():
    detected.append('Claude Code')
    primary_instruction_file = 'CLAUDE.md'
if (cwd / '.cursor').exists():
    detected.append('Cursor')
    ...

Detected environments: Claude Code (.claude/), Cursor (.cursor/), Aider (.aider/), Continue (.continue/)

bootstrap()

Orchestrates the full initialization sequence:

1. Calls detect_llm_environment() and reports findings.
2. Creates ouro/wiki/ and subdirectories (entities/, decisions/, patterns/, maps/) if they don't already exist.
3. Copies template files (index.md, schema.md, capture-queue.md) from the skill's own wiki/ directory.
4. Iterates over known instruction filenames and appends the maintenance protocol to each one found.
5. If no instruction file is found, creates one using the detected primary filename (fallback: AI_INSTRUCTIONS.md).

Instruction files checked: CLAUDE.md, GEMINI.md, CURSOR.md, CLINE.md, AIDER.md, CONTINUE.md, AI_INSTRUCTIONS.md

The script is idempotent: it skips wiki creation if ouro/wiki/ already exists, and skips protocol append if "Ourobor OS Maintenance Protocol" is already present in the file.

CLI Usage

# Run from the target project root after installing the skill
python <path-to-skill>/scripts/bootstrap.py

Post-Bootstrap Next Steps

1. Read ouro/wiki/index.md to verify initialization.
2. Run python ouro/scripts/capture.py --crawl to stage the existing codebase.
3. Ask the LLM to synthesize the capture queue into structured wiki entries.