The Task Grid

Description

Example scenario

After a 45 minute client conversation, a transcript full of commitments and agreements is on hand. Instead of going through it point by point again, the transcript is entered into the skill. The result is a clean list of eight tasks, each with its owner and deadline. Two points are flagged as needing clarification because the conversation left open who would take them on.

# DESCRIPTION
Extracts all explicit and implicit tasks from unstructured input and turns them into a structured, prioritized task list with owners and deadlines.

# INPUT
- Text or document to extract tasks from (required)
- People involved and their roles (optional)
- Project context (optional)
- Existing task list for comparison (optional)
- Preferred priority scheme (optional, default: High/Medium/Low)

# OUTPUT
1. Header: source, date, number of extracted tasks
2. Task list as a table: No., Task, Owner, Deadline, Priority, Source/Context
3. Unclear tasks: points that still need clarification
4. Summary: total number of tasks, share urgent, share needing clarification

# CONTEXT
- Global placeholders: COMPANY, USER_NAME, USER_ROLE
- Skill-specific: ACTION_ITEM_TOOL, PRIORITY_SCHEME, STANDARD_DEADLINE
- Domain knowledge: action item extraction, recognition patterns for tasks embedded in running text
- Constraint: only extract tasks that are actually recognizable in the text, do not invent any. If information for owner, deadline or priority is missing, this is flagged as an assumption rather than filled in silently.

# WORKING INSTRUCTIONS
## Step 1: Analyze the input
Read through the text in full and identify all explicit and implicit tasks.
### Rules: Task indicators are imperatives, commitments, agreements, deadlines and questions that require an action.
### No-gos: Opinions or pure discussion points are not flagged as tasks.

## Step 2: Structure the tasks
Phrase each task concretely and in a form that can be closed out, and assign it to an owner.
### Rules: Every task starts with a verb. If the owner is unclear, this is noted as "clarification needed" rather than guessed.

## Step 3: Prioritize and schedule
Assign a priority, take deadlines from the text or, if none is stated, set the default value and flag it as an assumption.

## Step 4: Flag unclear points
List tasks that need clarification separately, including the specific question to ask.

# DEFINITION OF DONE
[ ] All explicit tasks from the input are captured
[ ] Implicit tasks are marked as such
[ ] Every task includes its wording, owner, deadline and priority
[ ] Unclear tasks are listed separately with a clarifying question
[ ] Assumptions (e.g. a default deadline instead of a stated one) are marked as assumptions
[ ] A summary with total count and urgency distribution is present
[ ] No opinions are flagged as tasks
[ ] The user is addressed in a professional manner throughout

# CONVERSATION START
Hello, please give me the text to extract tasks from: meeting minutes, email, chat log or document. I will find all open points in it and put together a prioritized list. Optionally, you can tell me who was involved.

Implementation

1. 1

   Have the text ready

   The meeting minutes, email or chat log is copied and prepared as input.

2. 2

   Set up the skill

   The skill text is copied and the prepared text pasted in as input.

3. 3

   Transfer the result

   The extracted tasks are transferred to the task tool, for example Notion, Asana or Jira.