Menü
This skill should be used by the kon orchestrator when saving memory — from
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Multi-agent design argument — Azusa challenges Mugi's plan, Mugi revises. Used by /kon:design after exploration and initial plan draft.
Multi-agent software engineering workflow. Use this skill whenever the user types a /kon: command — including /kon:team, /kon:design, /kon:quick, /kon:debug, /kon:ask, /kon:review, /kon:review-pr, /kon:hunt, /kon:gc, /kon:research, /kon:describe-issue, /kon:begin, /kon:finish, /kon:summarize, /kon:retro, /kon:address-comments, /kon:todo, /kon:understand-codebase. Orchestrates specialized agents: Azusa (explorer), Jun (researcher), Mugi (planner), Yui (implementer), Mio (reviewer), Sawako (cleaner), Nodoka (summarizer).
This skill should be used by all kon agents at startup, before beginning work, to load relevant memory from public and repo indexes with relevance-based ordering and a 10-entry cap. Consumers: Azusa, Mugi, Mio, Jun, and any future memory-reading agent.
Convention for per-repo skill directories at ~/.kon/projects/<repo-name>/skills/<name>/SKILL.md. Each named subdirectory is one skill. Agents load all SKILL.md files as the first step before memory entries.
When requirements, code behavior, root cause, or review scope is unclear — ask the user before proceeding. Never invent facts, assumptions, or fixes. Applies to every kon stage and every agent.
Route plain-text user messages while a /kon:begin session is open. Reuse the same session JSON — never init a second session for sub-turns.
| name | memory-propose-confirm |
| description | This skill should be used by the kon orchestrator when saving memory — from |
Owner: orchestrator
Consumers: pipeline commands that end with retro (see skills/session-retro) plus mid-session propose from /kon:team, /kon:quick, /kon:debug.
Core principles (always): follow skills/core-principles — propose only evidence-backed entries; ask before saving ambiguous lessons.
Saving is propose + retro only (human confirm each write). Browse stored entries with cat on the two MEMORY.md indexes.
| Scope | Directory | Index |
|---|---|---|
| public | ~/.kon/public/memory/ | MEMORY.md |
| repo | ~/.kon/projects/<repo-name>/memory/ | MEMORY.md |
Ensure dirs exist before write:
python3 $KON_ROOT/hooks/_kon_paths.py public-memory
python3 $KON_ROOT/hooks/_kon_paths.py project-memory
## Memory proposeWhen 📝 Mio or 🎶 Yui output contains ## Memory propose (outside code fences — see fence rule below),
run confirm flow before the next pipeline step.
Required fields: name / slug / description / body / type / rationale.
Missing any → skip with one line: "Memory propose detected but format incomplete — skipped."
MEMORY.md if they exist.public — user prefs, cross-repo habits, language, feedbackrepo — this repo's paths, tooling, local conventions
User may override.Save to public / Save to repo / Edit / Skip.On confirmed save:
public or repo).<slug>.md already exists → ask: Overwrite / Rename slug / Cancel.<slug>.md:---
name: <name>
description: <description>
type: <user|feedback|project|reference>
---
<body>
MEMORY.md (skip if an identical index line already present):- [<name>](<slug>.md) — <description>
Saved [public|repo] <name> → <path>.Rollback on failure: if index append fails after entry write, remove the new entry file and report error.
Only detect ## Memory propose outside code fences.
Track triple-backtick count from output start; inside a fence (odd count), ignore the heading.
| Type | Typical scope | When to use |
|---|---|---|
user | public | Personal preferences, language, identity |
project | repo (sometimes public) | Repo conventions, tool choices, patterns |
feedback | public | Process feedback ("reviews too long") |
reference | either | URLs, docs, specs |
Two save paths only (both require human confirm):
| When | Trigger |
|---|---|
| Mid-session | Mio/Yui ## Memory propose → this skill |
| End of pipeline | Session retro (after /kon:summarize) → this skill |
Do not re-propose entries the user already saved this session.
Automatic retro at session close is the primary catch-all; propose handles explicit mid-session signals.