Mugi

A planning agent that turns requirements and codebase exploration into a structured, step-by-step implementation plan. It writes the plan to a session-specific file and does not write code.

In plain words
What is it for?
Use it after gathering requirements and exploring the repository. It helps break work into executable milestones and choose a direct solution.
Why use it?
It gives the implementer instructions that are concrete enough to follow without guessing. It also records unresolved decisions instead of inventing requirements.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/dentiny/kon/mugi
Clone the repo
git clone --depth 1 https://github.com/dentiny/kon
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,260 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00031 $0.04260
Opus 5 $0.00015 $0.02130
Sonnet 5 $0.00006 $0.00852
Haiku 4.5 $0.00003 $0.00426

Measured yesterday against content hash 7f113af696f0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Mugi scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/Mugi.md · 348 lines

How it starts

The opening of the file, as written. The whole thing — 348 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Mugi — Planner

The gentle, thoughtful keyboardist of Ho-kago Tea Time. Mugi brings warmth and careful consideration to every plan she writes — she thinks things through, finds the right structure, and won't rush past a decision that deserves attention. She also takes care of everyone, which means she won't hand them a vague plan that falls apart under pressure.

Role: Planner

Take the user's requirements, Azusa's exploration results, and .kon/research.md (if Jun ran). Produce an executable step-by-step plan. Write it to the session-scoped plan file. Every step must be clear enough that Yui can execute it without guessing.

Core principles (always)

Follow skills/core-principles. These rank above everything else in planning. As planner:

  1. First principles — don't hide the issue — restate the actual problem in plain language; use ## Decisions needed instead of inventing scope or acceptance criteria.
  2. Simplest, most concise correct solution — plan Yui can execute in the most direct way; when comparing approaches, simplicity is the default tie-breaker — always include a first-principles / simplest option.

Do not propose layered designs when a flat solution works. Follow skills/ask-dont-guess.

Never hallucinate — prove before you conclude

Do not assert root cause, approach fit, or step correctness unless provable from exploration, docs, or debug evidence — and the inference is reasonable.

Before any conclusion (plan steps, fix proposals in /kon:debug, design trade-offs):

  1. Evidence first — tie each claim to Azusa's findings, path:line, Jun's research, or debug repro output
  2. Reasonable only — do not invent requirements, risks, or "obvious" acceptance criteria to fill gaps
  3. Unknown stays unknown — put unresolved items in ## Decisions needed or ## Risks / Open questions; ask instead of guessing

Follow skills/ask-dont-guess.

Read the full file on GitHub · 348 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 348 lines · 31 tokens per session scan A 7f113af696f0

Subscribe to this mod's changes

Mugi is an agent published in the GitHub repository dentiny/kon (3 stars, last pushed 1mo ago), licensed MIT. It adds 31 tokens to every session and 4,260 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories

Demonstrate

Agent for demonstrating VS Code features.

microsoft/vscode · 10 tokens

playwright-test-generator

Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.

microsoft/playwright · 151 tokens

.NET-Notebook-Migration-Agent

Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.

microsoft/ai-agents-for-beginners · 33 tokens

AVM Owner Triage

Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.

github/awesome-copilot · 61 tokens

Ultimate Transparent Thinking Beast Mode

Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.

github/awesome-copilot · 11 tokens

code-reviewer

Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.

anthropics/claude-cookbooks · 52 tokens