{
  "markdown": "<p align=\"center\">\n  <img src=\"public/banner.svg\" alt=\"Ayanokoji — Strategic Execution Protocol\" width=\"1080\"/>\n</p>\n\n# Ayanokoji — Strategic Execution Protocol\n\nA universal strategic execution protocol for AI coding agents. It replaces reactive, error-prone behaviors with pre-calculated, structured planning and verification. Works across **Claude Code, Cursor, Windsurf, GitHub Copilot, Cline, Kiro, OpenCode, Hermes**, and any LLM agent.\n\n---\n\n## Key Features\n\n* **Methodical Intake**: The agent silently analyzes tasks and asks at most 9 targeted questions to define the end state, hard constraints, and acceptance criteria before writing any code.\n* **Persistent Memory Scaffold**: Generates four root-level markdown files to maintain state across resets:\n  * `AGENTS.md` — Rulebook (hard-capped at 60 lines) with constraints and operating principles.\n  * `CONTEXT.md` — High-level architecture map and stack definition.\n  * `TASK.md` — Bounded mission brief (files in scope, approach, risks, and criteria). Overwritten per task.\n  * `PROGRESS.md` — Intelligence log of completions (with proof), failures (with exact errors), and blockers.\n* **Mandatory Verification Gate**: Enforces execution of TASK.md's exact acceptance criteria. \"Tests passed\" is not allowed; concrete command outputs or detailed checks must be logged in PROGRESS.md.\n* **Fluff & Slop Control**: Restricts LLM outputs to direct code, plan updates, and answers, forbidding greetings (hi, hello) or verbose explanations of what the code does.\n\n---\n\n## Workflow\n\n<p align=\"center\">\n  <img src=\"public/workflow.png\" alt=\"Ayanokoji Workflow\" width=\"1080\"/>\n</p>\n\n---\n\n## Commands\n\nAll commands are available as slash commands, `@mentions`, or natural language directives.\n\n| Command | Action |\n|---|---|\n| `/ayanokoji` | Activate the strategic execution protocol for this project. |\n| `/ayanokoji-task [desc]` | Overwrite and initialize `TASK.md` with the new task description. |\n| `/ayanokoji-gate` | Verify the current task's acceptance criteria and update scaffold logs. |\n| `/ayanokoji-status` | Display the current task summary, blockers, failures, and scaffold health. |\n| `/ayanokoji-review` | Inspect the active `TASK.md` for completeness and risks before starting. |\n| `/ayanokoji-audit` | Audit all four scaffold files for protocol compliance and syntax errors. |\n| `/ayanokoji-debt` | List all deferred decisions, `[TO BE DEFINED]` blocks, and blockers. |\n| `/ayanokoji-gain` | Summarize project stats: completed tasks, retry rate, and verification quality. |\n| `/ayanokoji-help` | Display the protocol quick-reference guide. |\n\n*Deactivate by saying `\"stop ayanokoji\"` or `\"normal mode\"`.*\n\n## Setup Guide\n\n### Instant Setup (Recommended)\nAutomatically configure the Ayanokoji ruleset for all detected IDEs in your current project directory with a single command:\n```bash\nnpx ayanokoji.md\n```\nThis command writes the master `AGENTS.md` to your project root and configures the relevant rule files for Cursor, Windsurf, Copilot, Cline, or Kiro based on your project structure.\n\n---\n\n## Manual IDE Setup Guide\n\n### 1. Cursor\nCopy the cursor rule to your project:\n```bash\nmkdir -p .cursor/rules\ncp .cursor/rules/ayanokoji.mdc .cursor/rules/ayanokoji.mdc\n```\nThe rule will automatically steer all composer and chat sessions.\n\n### 2. Windsurf\nCopy the windsurf rule to your project:\n```bash\nmkdir -p .windsurf/rules\ncp .windsurf/rules/ayanokoji.md .windsurf/rules/ayanokoji.md\n```\n\n### 3. Claude Code / Gemini CLI / Antigravity CLI\nInstall as a plugin:\n```bash\nclaude mcp add ayanokoji https://github.com/Sahnik0/ayanokoji.md\n```\nOr use the native commands configured in `commands/` and `gemini-extension.json`.\n\n### 4. GitHub Copilot\n* **Instruction-tier**: Copy `.github/copilot-instructions.md` to your repository.\n* **Plugin-tier**: Add it through the Copilot Plugin Marketplace.\n\n### 5. Cline\nCopy the rules file to your project:\n```bash\ncp .clinerules/ayanokoji.md .clinerules\n```\n\n### 6. Kiro\nCopy the steering rule:\n```bash\nmkdir -p .kiro/steering\ncp .kiro/steering/ayanokoji.md .kiro/steering/ayanokoji.md\n```\n\n### 7. OpenCode\nAdd the plugin dependency to your `opencode.json`:\n```json\n{\n  \"plugins\": [\"ayanokoji.md\"]\n}\n```\n\n### 8. Hermes Agent\nInstall and enable the plugin:\n```bash\nhermes plugins install ayanokoji\nhermes plugins enable ayanokoji\n```\n\n---\n\n## Project Structure\n\n```\nayanokoji.md/\n  ├── AGENTS.md                          — Canonical ruleset (copied by agents)\n  ├── skills/                            — Full protocol markdown files\n  │   ├── ayanokoji/SKILL.md             — Base strategic protocol\n  │   └── ayanokoji-[sub]/SKILL.md       — Sub-skill rules (review, status, etc.)\n  ├── commands/                          — Slash command descriptors (.toml)\n  ├── hooks/                             — Adapter hooks & scripts for shells/CLIs\n  ├── pi-extension/                      — Personal AI (pi) extension files\n  ├── scripts/                           — Development lint, build, & sync scripts\n  ├── tests/                             — Complete functional unit test suite\n  ├── benchmarks/                        — Correctness and LOC graders\n  ├── __init__.py                        — Hermes plugin runtime entry\n  └── plugin.yaml                        — Hermes plugin manifest\n```\n\n---\n\n*\"Emotions are inefficiencies. I calculate, therefore I execute.\"*\n",
  "bytes": 5371,
  "sha": "cbff5ee6bed1325582b49d629afaebc5425beebeeb790050aaa586a420adf22c",
  "repo_slug": "atanu2k4/ayanokoji.md",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_atanu2k4_ayanokoji_md_2b87d4cf/readme"
}