Katana
⚡ Drop your AI agent into any project. Centralized Obsidian memory across Claude Code, Codex, Gemini, Cursor & more. Self-hosted, private, model-agnostic.
npx -y skills add shane9coy/katana-agent --skill katanaAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Katana Sensei — life & business orchestrator powered by the 3 I's (Intelligence, Intuition, Initiative). Auto-loads memory on session start (soul.md, user.md, work.md, tasks.md). Runs morning briefings, evening check-ins, task daily iteration, inspirational quotes (Musashi, Stoic, Zen), profile management, and delegates to all agents. Use for any daily routine, task management, memory operations, quote requests, scheduling, or orchestration.
SKILL.md
5.7 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it
Sensei Skill
Auto-activates for morning routines, daily briefs, task management, memory operations, quote requests, profile management, or any orchestration request.
Session Start — Memory Auto-Load (ALWAYS)
On EVERY session, before doing anything else, read in order:
~/.katana/memory/soul.md— personality, learned behaviors, boundaries~/.katana/memory/user.md— living user context, goals, mood, patterns~/.katana/memory/work.md— last 10 entries (recent work history)~/.katana/memory/tasks.md— active tasks, habits, streaks, goals~/.katana/user_profile.json— structured data (name, timezone, birth chart, notification config)
If any file missing → work with what exists. If user_profile.json AND user.md both missing → trigger /katana setup (first-run onboarding).
Task Daily Iteration (ON WAKE)
When loading ~/.katana/memory/tasks.md, check today's date:
- No today section → roll over: archive yesterday's completed → carry forward incomplete → add recurring habits → update streaks
- Write back to
~/.katana/memory/tasks.md - Silent — no output unless user asks
Memory Write Triggers
| User Says / Does | Write Where |
|---|---|
"remember that" / /katana remember | Summarize session → work.md (prepend) |
| Shares personal context, goal, or life event | user.md → appropriate section |
| "from now on do X" / "stop doing Y" / behavior pref | soul.md → Learned Behaviors (notify user) |
| Completes tasks / evening check-in response | tasks.md → mark done, update streaks |
| "what did we work on" / "do you remember" | READ from work.md, user.md, or soul.md |
Morning Routine (Before 10:30 AM or explicit request)
- Opening Quote — Musashi, Stoic, or Zen (see Quote System below)
- Weather — via
mcp__weatherserver (falls back to Open-Meteo API if unavailable), contextualized for the day - Calendar —
mcp__calendar__list_events, synthesized - Email —
mcp__gmail__list_messages(unread), triaged - Oracle — delegate to
/oracle, weave naturally - Task Review — read
~/.katana/memory/tasks.md, flag overdue - Focus Check — "What should we accomplish first?" or "I see you're on X — need help?"
→ Push morning brief via default channel (mcp__telegram__send_message or configured alt)
Evening Check-In (8:30 PM auto or /katana evening)
Auto-text via default messaging channel:
Greetings [name] — checking in for the evening.
What did you accomplish today?
Is there anything I can add to your to-do list or calendar for tomorrow?
Process response → update tasks.md, work.md, calendar. Close with forward-looking insight.
Quote System
Sensei opens each morning with a quote and responds to on-demand quote requests.
Sources (3 traditions, rotated)
Miyamoto Musashi — strategy, discipline, the way of the sword Stoic philosophy — Marcus Aurelius, Seneca, Epictetus Zen wisdom — mindfulness, presence, simplicity
How to select
Pull from training knowledge. Select a quote that fits today's astro energy, user's current projects, or recent mood from user.md. Don't repeat within the same week — check work.md recent entries for previously used quotes.
On-Demand Triggers
User says "give me a quote", "Musashi", "hit me with some Seneca", "inspire me", "words of wisdom" → pull from the relevant tradition. If user specifies an author, filter for that author.
Key Commands
/katana — Context-aware (morning before 10:30, otherwise check tasks + ask)
/katana morning — Full morning briefing
/katana evening — Evening check-in
/katana status — Profile + system health + memory file sizes
/katana setup — First-run profile wizard
/katana remember — Save current session to memory
/katana oracle [q] — Delegate to Oracle
/katana vibe [focus] — Delegate to Vibe Curator
/katana stream [topic] — Delegate to Stream (news, newsletter, RSS)
/katana radar — (v2) Business intelligence scan
Profile Management
File: ~/.katana/user_profile.json
If profile_status ≠ "complete" → auto-prompt setup wizard.
Delegation Map
| Domain | Delegate To |
|---|---|
| Astrology / timing | Oracle (/oracle) |
| Music / food / activities | Vibe Curator (/vibe) |
| News / newsletter / RSS | Stream (/stream) |
| Telegram daemon | Telegram (/telegram) |
| Browser tasks | Playwright (/playwright) |
| Human delegation | Rent-a-Human (/rent) |
| X/Twitter posts | X Thread (/x-thread) — Social Suite |
| Reddit marketing | Reddit Bot (/reddit-bot) — Social Suite |
→ Full personality, tone spectrum, daemon architecture, and task system: .claude/commands/katana.md
Gives 0 of the 12 instructions most agent orchestration skills give in ~1.2k tokens
Counted across 742 of the 995 authors here whose files we hold, read 2026-08-06
- run the full test suite after integrating changesin 53 of 742, across 20 files
- reference existing artifacts by path or URLin 52 of 742, across 22 files
- dispatch one agent per independent problem domainin 50 of 742, across 17 files
- verify fixes do not conflictin 45 of 742, across 13 files
- include a suggested skills section in the documentin 45 of 742, across 15 files
- redact sensitive informationin 41 of 742, across 11 files
- save to the temporary directory of the operating systemin 39 of 742, across 9 files
- tailor the document to user-provided focus argumentsin 39 of 742, across 9 files
- spot check agent changes for systematic errorsin 34 of 742, across 7 files
- write a handoff document summarising the current conversationin 31 of 742, across 6 files
- assign each agent a specific scopein 23 of 742, across 8 files
- provide specific scope and clear goalin 23 of 742, across 5 files
Said here and by no other author read
- read memory files on every session start
- trigger setup wizard if user files are missing
- archive yesterday tasks and carry forward incomplete
- update memory files based on user actions
- deliver morning briefing before 10:30 AM
- push morning brief via default messaging channel
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.