OIN-SMILES

Updates all project documentation after code or skill changes. Always reads actual implementation before writing. Covers README, CHANGELOG, docs/, AGENTS.md, memory files, and skills-info.md. Use after completing any feature, fix, refactor, or skill update.

Compacts the current conversation into a handoff document so a fresh agent can continue the work. Saves to the OS temp directory, not the workspace. Pass an optional argument describing what the next session will focus on.

Documents the process, methodology, and decisions of the current session as a reproducible narrative. Saves to spec/process/ in the workspace so the journey can be understood and recreated by others.

Scaffolds a new project by conducting an interview to collect metadata and generating standard repository documentation files from the framework templates. Use when setting up a new repository.

Run this after /hyper-tutorial-run to turn tutorial failures into a structured, constraint-aware goal prompt. Interviews the user about their goal and what may/may not be changed, then generates a paste-ready prompt for a new Claude Code fix session. Does NOT re-run tutorial commands.

Generates a markdown tutorial from an integration test or user-provided files through iterative LLM + human collaboration. Tutorials are saved to tutorials/<name>/ with organized supporting files.

Interactively walks the user through a tutorial folder generated by /hyper-tutorial-generator. Reads tutorial.md section by section, answers questions in context, and surfaces linked code samples and output files on demand.

Enhanced commit-push-PR automation: AI-suggests commit messages from CHANGELOG, uses HITL approval gates at each step, optionally creates PRs with auto-drafted descriptions. No system-level git/gh approval prompts required.

Relentlessly interviews the user one question at a time, providing a recommended answer with each question and exploring the codebase before asking anything derivable from existing code. Produces a Draft PRD.

Strictly verifies the codebase against a specific MiniPRD and reconciles the Hypergraph memory.

A prompt designed to instantiate a peer-level AI research partner for highly complex technical fields like software engineering, drug discovery, and materials science. It actively challenges assumptions using First Principles thinking, enforces strict anti-hallucination rules with standardized citations, and maintains an exhaustively detailed "Living Master Plan" (including a "Graveyard" of failed iterations) to track long-term project evolution.

Activates a CTO persona to collaboratively plan and review technical decisions before entering the formal Hypergraph specification pipeline. Use when brainstorming architecture, evaluating trade-offs, or planning a new feature with a trusted technical advisor.

Quickly captures a bug, feature request, or improvement as a structured GitHub issue without interrupting the user's development flow. Use when the user needs to log something fast while staying focused on current work.

Temporarily pauses the main development plan to exhaustively explore, research, or brainstorm a specific topic using First Principles thinking and strict epistemic humility. Use when the user needs to deeply understand a concept before making a decision.

Implements the current plan precisely and in full, following existing code patterns and updating the hypergraph after every modification. Use when ready to write code against a compiled MiniPRD.

Relentless domain-sharpening session that challenges your plan against the existing domain model, enforces precise terminology, and updates CONTEXT.md and ADRs inline as decisions crystallise. Use when stress-testing a plan against the project's language and documented decisions.

Pauses development mode to teach a technical concept at three increasing levels of complexity, tailored to a technical PM audience. Use when the user wants to understand something they encountered while building.

Converts a desired behavior, prompt, or idea into a properly formatted workflow file in the Hypergraph Coding Agent Framework. Use when the user wants to create a new slash command or workflow.

Evaluates external peer review findings against the actual codebase, separating valid issues from misunderstandings, and produces a prioritized action plan. Use when another model or reviewer has provided feedback on the current implementation.

Collaboratively designs an optimal, personalized prompt with the user using advanced LLM prompt engineering best practices. Explains the rationale behind structural choices educationally. Framework-agnostic. Use when the user wants to create or refine an AI prompt.

Mediates Red Team findings, forces architectural trade-offs, and compiles the final SuperPRD and MiniPRDs.

Explains the Master SOP and guides the user to the correct framework phase. Use when a new user needs orientation or when you need to re-anchor to the correct workflow phase.

Translates rough UI/UX ideas and visual references into a concrete Design System specification with tokens, component specs, and layout rules. Use when starting UI work or defining visual standards for a feature.

Reverse-engineers a filled-out source document into a reusable Markdown template for a downstream AI agent. Use when the user provides a completed document and wants to extract a reusable template from it.

Guides the user through diagnosing and recovering from common failure states in the Hypergraph Coding Agent Framework. Use when the user reports bugs, hallucinations, desynchronization, or errors with the framework agents.

Walks the user through a practical, step-by-step example of the Hypergraph Coding Agent Framework using the Email Newsletter Subscription scenario. Use when onboarding new users or demonstrating the framework workflow.

Interactive smart-upgrade command. Fetches latest upstream framework, auto-updates non-sensitive files, and guides user through section-by-section collaborative merge of sensitive customizable files. Backs up any replaced or merged file to .agents/.backup/YYYY-MM-DD/.

Audits and fixes HACF agent instruction files (CLAUDE.md, AGENTS.md, GEMINI.md) in an installed project to ensure HACF is framed as a development toolchain, not the project subject. Use when HACF content bleeds into project plans, PRDs, or architecture docs.

Reverse-engineers the existing codebase to generate a "Current State" SuperPRD.

Flushes conversation history and resets session context between feature cycles. Idempotent—safe to call multiple times.

Converts an agentic prompt or Gemini Gem into a new structured skill for the Hypergraph Coding Agent Framework. Use this skill when asked to create a new skill from a prompt, gem, or idea.

Scans the existing codebase to initialize or update the architecture.yml hypergraph.

Performs an adversarial Blast Radius and vulnerability analysis on the Draft PRD.

Reads all memory, rules, docs, and the codebase to rebuild a fresh mental model of the project, then updates the memory files to reflect actual current state. Use when starting a new session or when memory feels stale.

Updates agentic memory files (activeContext, systemPatterns, productContext) and project docs with work done in the current session. Use at the end of a development session to keep memory in sync with the codebase.

Immediately outputs the current, fully updated Living Master Plan including project objectives, architecture hypotheses, actionable steps, constraints, and a graveyard of failed approaches. Use at any point to get a snapshot of the project state.

Executes a state-machine interview to extract exhaustive requirements and generate a Draft PRD.

Strictly verifies the codebase against a specific MiniPRD and reconciles the Hypergraph memory.

Reverse-engineers the existing codebase to generate a "Current State" SuperPRD.

A prompt designed to instantiate a peer-level AI research partner for highly complex technical fields like software engineering, drug discovery, and materials science. It actively challenges assumptions using First Principles thinking, enforces strict anti-hallucination rules with standardized citations, and maintains an exhaustively detailed "Living Master Plan" (including a "Graveyard" of failed iterations) to track long-term project evolution.