sub-tasks
Skill9 starsChanged 8 months agoMITView on GitHub
Manage the project's Sub-Task system. Sub-Tasks are SubFrame's project task tracking stored as individual markdown files in .subframe/tasks/. Current task index: !cat .subframe/tasks.json 2>/dev/null || echo "No tasks.json found" Argument: $ARGUMENTS Each task is a markdown file in .subframe/tasks/ with YAML frontmatter: Regenerate the .subframe/tasks.json index by reading all .subframe/tasks/*.md files (excluding archive/) and building the JSON structure with tasks grouped by status (pending, inProgress, completed). Show the current task list and ask the user what they'd like to do:…
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 4 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 58
Read .subframe/tasks.json for the index overview, or glob .subframe/tasks/*.md and read frontmatter.
- Very specific: line 61
Read the specific .subframe/tasks/<id>.md file.
- Very specific: line 70
Create a new .subframe/tasks/<id>.md file with:
- Distinctive: line 89
Start a pending task
- Distinctive: line 91
Create a new task
- Distinctive: line 100
When the user says things like "let's do this later", "add a task for...", or "we should...":
- Very specific: line 102
Write a detailed description explaining what, how, and which files
- Distinctive: line 104
Create the .md file
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- @subframe-version 0.14.0-beta --><!-- @subframe-managed -->name: sub-tasks description: View and manage SubFrame Sub-Tasks. Use when starting work, completing tasks, checking what's pending, or creating new tasks from conversation. disable-model-invocation: false argument-hint: [list|start|complete|add|get|archive] allowed-tools: Bash, Read, Write, Edit, Glob
SubFrame Sub-Tasks
Manage the project's Sub-Task system. Sub-Tasks are SubFrame's project task tracking stored as individual markdown files in .subframe/tasks/.
Dynamic Context
Current task index:
!cat .subframe/tasks.json 2>/dev/null || echo "No tasks.json found"
Instructions
Argument: $ARGUMENTS
Task File Format
Each task is a markdown file in .subframe/tasks/ with YAML frontmatter:
---
id: task-abc123
title: My task title
status: pending
priority: medium
category: feature
description: What needs to be done
userRequest: The user's original words
acceptanceCriteria: How to verify completion
blockedBy: []
blocks: []
createdAt: 2024-01-01T00:00:00.000Z
updatedAt: 2024-01-01T00:00:00.000Z
completedAt: null
---
## Notes
Session notes go here.
## Steps
- [ ] Step one
- [x] Step two (completed)Operations
List tasks
Read .subframe/tasks.json for the index overview, or glob .subframe/tasks/*.md and read frontmatter.
Get task details
Read the specific .subframe/tasks/<id>.md file.
Start a task (pending → in_progress)
Edit the task's frontmatter: set status: in_progress and update updatedAt.
Complete a task
Edit the task's frontmatter: set status: completed, set completedAt to current ISO timestamp, update updatedAt.
Add a new task
Create a new .subframe/tasks/<id>.md file with:
- Generate id:
task-+ 8 random alphanumeric chars - Set
status: pending,createdAtandupdatedAtto current ISO timestamp completedAt: null- Include all required fields in frontmatter
Update a task
Edit the frontmatter fields as needed. Always update updatedAt.
Archive completed tasks
Move completed .md files to .subframe/tasks/archive/YYYY/ (create directory if needed).
After Any Write Operation
Regenerate the .subframe/tasks.json index by reading all .subframe/tasks/*.md files (excluding archive/) and building the JSON structure with tasks grouped by status (pending, inProgress, completed).
If invoked without arguments
Show the current task list and ask the user what they'd like to do:
- Start a pending task
- Complete an in-progress task
- Create a new task
- Archive completed tasks
If invoked with a task ID
Show full details for that task by reading its .md file.
Creating tasks from conversation
When the user says things like "let's do this later", "add a task for...", or "we should...":
- Capture the user's exact words as
userRequest - Write a detailed
descriptionexplaining what, how, and which files - Set appropriate
priorityandcategory - Create the .md file
- Regenerate the index
- Confirm the task was created
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

