Weekly Planning logo

Weekly Planning

Community
serejaris
weekly-planning

Use when the user is transitioning from a completed retro into a weekly plan, choosing weekly outcomes, scheduling a full ISO week, or asking for "план на неделю", "weekly planning", priorities, capacity, or outcomes.

Overview

Publisherserejaris
Repositorypersonal-corp-os
Skill nameweekly-planning
Stars
232
Forks
26
Bundled files
4
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 4 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by serejaris on GitHub. Read the source before you install it.

Installation

Install the Weekly Planning AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/serejaris/personal-corp-os/tree/main/skills/weekly-planning
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/serejaris/personal-corp-os.git /tmp/personal-corp-os
mkdir -p .claude/skills
cp -r /tmp/personal-corp-os/skills/weekly-planning .claude/skills/weekly-planning
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Weekly Planning in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Weekly Planning on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Weekly Planning is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Weekly Planning

Overview

Turn retro findings + existing backlog into a prioritized week with clear outcomes, issues in correct repos, and delegation matrix (founder vs agent).

When to Use

  • After weekly retro is complete
  • Monday/Tuesday when planning the week
  • When user says "план на неделю", "приоритизация", "что делаем"

Iron Rule: Planning Is Not Execution

Weekly planning creates a plan. It does not execute the plan.

Allowed during planning:

  • Gather context needed to choose outcomes
  • Read calendars, issues, project boards, CRM/backlog files, and recent notes
  • Draft outcomes, issue proposals, and board updates

Not allowed during planning:

  • Start implementation work for a chosen outcome
  • Create deliverables, briefs, proposals, content, code, or production changes
  • Spawn executor agents for work that belongs to the coming week

If the user asks to "do this now" while planning, finish or pause planning first, then handle that as a separate execution task.

Process

dot
digraph planning {
  rankdir=TB;
  node [shape=box];

  collect [label="1. Collect all tasks"];
  calendar [label="2. Map calendar"];
  group [label="3. Group by surface"];
  eisenhower [label="4. Eisenhower split"];
  outcomes [label="5. Choose outcomes"];
  issues [label="6. Create issues\nin correct repos"];
  board [label="7. Add to board\nwith W{NN} label"];
  page [label="8. Create living\nweekly plan"];

  collect -> calendar -> group -> eisenhower -> outcomes -> issues -> board -> page;
}

1. Collect all tasks

Sources to scan:

  • Retro backlog (label retro:W{NN-1})
  • Open issues across ALL repos: gh search issues --owner $YOUR_OWNER --state open
  • GitHub Projects for active initiatives
  • Calendar events (meetings, lectures, mentoring)
  • User input (new tasks from interview)

2. Map calendar

Read known sources before asking the user. Use calendar, existing issues, project boards, and planning files to establish:

  • Today's date
  • Planned week start/end
  • Fixed events
  • Known deadlines
  • Available capacity

Ask only for facts that are not recoverable from sources, one question at a time.

IMPORTANT: Verify dates before writing outcomes. Do not create deadlines in the past. If the deadline is today, include a time boundary such as "by EOD".

3. Group by surface

Group tasks by where work happens. Use the repos and task routing defined in AGENTS.md / CLAUDE.md config (set up by corp-init).

Example surfaces:

SurfaceDescription
Product deliveryMain product/service repos
Sales/pipelineCRM, sales tooling
ContentBlog, videos, social media
StrategyPlanning, cross-cutting decisions
InfrastructureInternal tools, automation

4. Eisenhower split with delegation

For each task, classify:

UrgentNot Urgent
ImportantFounder does TODAYFounder schedules this week
Not ImportantAgent does asyncParking lot / drop

Key question per task: "Does this require founder judgment, or can an agent execute it from the issue description?"

5. Choose outcomes

Outcomes = results, not tasks. Format: "By Friday, X is true."

Rules:

  • NOT limited to 3 — as many as realistic
  • Each outcome has a measurable check
  • Include both founder-only and agent-delegated work
  • Include B2B/sales if there are active deals — these are outcomes, not bonuses
  • Outcomes are commitments chosen with the user, not a silent reconstruction from the backlog
  • Use the backlog as input, then ask the user to choose or edit outcome options

Example:

Outcome: New feature launched
Check: deployed to production, 0 critical bugs by Friday
Founder: final UX decisions, announcement copy
Agent: implementation, tests, deploy script

6. Create issues in CORRECT repos

CRITICAL: Issue lives where the work happens. Use the task routing from your CLAUDE.md config.

Issue about backend bug → backend repo
Issue about marketing → marketing repo
Issue about content → content repo
Issue about strategy → main planning repo

Before creating any issue:

  1. VERIFY current state (check existing issues, files, data)
  2. PROPOSE to user
  3. CREATE only after verification

Label every issue: W{NN} + optionally retro:W{NN-1} if from retro.

7. Add to unified board

All W{NN} issues → your GitHub Project (from CLAUDE.md config: project_id and owner).

Create or use view "W{NN}" filtered by label.

bash
# Add to project (use project_id and owner from your config)
gh project item-add $PROJECT_ID --owner $YOUR_OWNER --url $ISSUE_URL

# Verify
gh search issues --owner $YOUR_OWNER --label "W{NN}" --state open

8. Create the living weekly plan

Resolve the ISO week from the current local date. Never hardcode W34 or copy the previous filename.

bash
python3 skills/weekly-planning/scripts/living_week.py context --workspace . --json
python3 skills/weekly-planning/scripts/living_week.py create --workspace .

The generator creates reports/WNN-plan.html. Before presenting it:

  1. Remove every element with data-placeholder="true".
  2. Populate all seven day columns with accepted work; every task card has one data-area.
  3. Keep exactly one pair of daily:updated, daily:kanban, and hidden daily:day-slice markers. The local daily skill may replace only these marked regions.
  4. Mark past days as closed with the tasks and facts that actually happened. Highlight the current day. Keep future days as plan.
  5. Configure the three display time zones and the seven work Areas in the page script. Card color comes from Area metadata.
  6. Put the goal captured during retro/planning into each Area definition. The Area Explorer derives cards from the weekly board; it never stores a second task list.
  7. Keep Wealth as the recovery Area or rename it in config. Its daily habit checkboxes use a week-scoped localStorage key and must not be overwritten by agents.

The HTML page is a mutable view. Issues/task records and WNN-outcomes.md remain sources of truth. Daily updates preserve the full ISO week and mutate only the marked regions.

Red Flags — STOP

  • Creating all issues in one repo → WRONG, route to correct repo per config
  • Creating issue without checking current state → VERIFY FIRST
  • Claiming status without checking actual data → VERIFY
  • Planning 40h of productive work → budget 60% capacity
  • Skipping calendar mapping → leads to conflicts
  • Mixing "task" with "outcome" → outcome = result by Friday
  • Doing execution work inside planning → finish planning first, execute separately
  • Asking the user for facts that are already available in sources → read sources first
  • Writing deadlines without checking the actual date → verify current date and week range
  • Replacing the entire plan during daily update → mutate only the marker regions and preserve browser habit state
  • Copying last week's filename or localStorage key → resolve the current ISO week with the generator

Common Mistakes

MistakeFix
All issues in one repoRoute to where work happens
Issue created without verificationCheck issues + files + data first
Outcomes = task listRewrite as "by Friday, X is true"
Ignoring B2B as "bonus"Active deals with money = outcomes
Publishing paid content publiclyPaid content stays in paid repos
Assuming data without checkingVerify with gh issue list + actual files
Fixed 3 outcomes limitAs many outcomes as realistic
Executing tasks during planningCapture as outcomes/issues, execute after planning
Asking instead of readingCheck source systems first, ask only for missing facts

Integration with weekly-retro

This skill runs AFTER weekly-retro. Expected inputs:

  • Retro summary with what worked / didn't
  • Issues created during retro (label retro:W{NN-1})
  • Lessons learned and playbook updates
  • New tasks from retro interview

Setup

This skill reads config from your project's AGENTS.md / CLAUDE.md. Run corp-doctor first to generate or repair the config block, or add manually:

markdown
## Agent Operations Config

### Repos (scanned during retro/planning)
repos:
  - owner/repo-1    # what it does
  - owner/repo-2    # what it does

### GitHub Project
project_id: N
owner: your-github-handle

### Task Routing
routing:
  - pattern: "backend, bugs"
    repo: owner/backend-repo
  - pattern: "content, marketing"
    repo: owner/content-repo
  - pattern: "strategy, cross-cutting"
    repo: owner/main-repo

### Living plan
reports_dir: reports
local_timezone: system
display_timezones:
  - Europe/Berlin
  - Europe/Moscow
  - America/Argentina/Buenos_Aires
areas:
  - delivery
  - sales
  - systems
  - content
  - people
  - strategy
  - wealth
wealth_habits:
  - sport
  - reflection

Quick Reference

bash
# All tasks for current week
gh search issues --owner $YOUR_OWNER --label "W{NN}" --state open

# By repo
gh search issues --owner $YOUR_OWNER --label "W{NN}" --state open --json repository,number,title

# Close week (Friday)
gh search issues --owner $YOUR_OWNER --label "W{NN}" --state open --json title,repository
# Review: done / spillover / drop

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Weekly Planning AI skill do?

Use when the user is transitioning from a completed retro into a weekly plan, choosing weekly outcomes, scheduling a full ISO week, or asking for "план на неделю", "weekly planning", priorities, capacity, or outcomes.

Why use Weekly Planning on TypingMind?

Because you install it once and use it with any model. Weekly Planning is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Weekly Planning in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/serejaris/personal-corp-os/tree/main/skills/weekly-planning. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Weekly Planning?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Weekly Planning?

As many as you like. As long as a model supports skills, you can use Weekly Planning with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Weekly Planning AI skill free?

Yes. It is published on GitHub by serejaris under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇