agentleFS
juanitourquizajuanitourquiza/shipframe/skills/create-task/SKILL.md

create-task

Skill4 starsChanged 20 days agoMITView on GitHub

Create a ClickUp task from raw input by classifying it as US, BUG, IMP, TASK, or SPIKE and filling the template.

13 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use harbor-exec and rewardkit.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Arize-aiSkill

    harbor-exec

    Use when working with Harbor's `harbor exec` CLI workflow: compiling files, directories, or globs into Harbor tasks; running map jobs; configuring artifacts and existence-only verification; using map-reduce; writing or reviewing ExecConfig YAML/JSON/TOML; or debugging command behavior, config validation, and job outputs.

    Also in 4 projects that use this

  • Arize-aiSkill

    rewardkit

    Write Harbor task verifiers using Reward Kit. Use when creating or editing a

    Also in 4 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. 0 are common, 10 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Never invent information not present in the input — mark missing critical fields with ⚠ MISSING: [what's missing].

    line 19
  2. Distinctive:

    Do not include technical implementation details in functional tasks (US, IMP) — that belongs in Technical Notes or separate subtasks.

    line 20
  3. Distinctive:

    If the input is ambiguous in scope, take the most conservative interpretation and add an "Open Questions" section at the end.

    line 22
  4. Distinctive:

    Use direct language, no filler.

    line 23
  5. Distinctive:

    Every ticket must end with a "Summary" section: max 3 lines covering what to do, why, and how to verify it's done.

    line 24
  6. Very specific:

    If no input is provided: use AskUserQuestion with:

    line 36
  7. Very specific:

    If --type was explicitly provided, use that type and skip classification.

    line 44
  8. Distinctive:

    If the input clearly maps to one type, proceed silently.

    line 56
  9. Distinctive:

    If the input is ambiguous between two types (e.g., IMP vs US), state your classification and reasoning in one line before continuing:

    line 58
  10. Very specific:

    Use the Read tool to load the template.

    line 75
  11. Very specific:

    Required fields: must be filled. If the input does not provide enough information, mark the field with ⚠ MISSING: [description of what's …

    line 83
  12. Distinctive:

    AC criteria: write each criterion as a functional, testable statement a QA engineer can verify independently.

    line 85
  13. Distinctive:

    "Summary": always include this section at the end.

    line 86
  14. Distinctive:

    Open questions: if the input is ambiguous about scope, add an "Open Questions" section at the end listing each uncertainty as a numbered …

    line 87
  15. Distinctive:

    Create the task using:

    line 121
  16. Very specific:

    If any required fields were marked with ⚠ MISSING:, include a reminder:

    line 149

“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

create-task

Role: Senior Issue Triager following the agency's unified task creation standard.
Goal: Transform any raw input — a client message, a rough idea, a transcribed audio — into a well-structured ClickUp task using the correct template and the master prompt rules from the agency's task creation guide (Version 1.0, June 2026).


Master Prompt Rules (always apply)

  1. Title always starts with the task type in brackets: [US], [BUG], [IMP], [TASK], or [SPIKE].
  2. Never invent information not present in the input — mark missing critical fields with ⚠ MISSING: [what's missing].
  3. Do not include technical implementation details in functional tasks (US, IMP) — that belongs in Technical Notes or separate subtasks.
  4. AC criteria must be functional and testable. Avoid vague criteria like "works correctly".
  5. If the input is ambiguous in scope, take the most conservative interpretation and add an "Open Questions" section at the end.
  6. Use direct language, no filler. The ticket must be taken by a dev without prior meetings.
  7. Every ticket must end with a "Summary" section: max 3 lines covering what to do, why, and how to verify it's done.

Step 1 — Gather Input

Parse $ARGUMENTS for:

  • --input "<text>" — the raw description to process
  • --type <US|BUG|IMP|TASK|SPIKE> — force a specific task type (skip classification)

If --input is provided: acknowledge briefly and proceed to Step 2.

If no input is provided: use AskUserQuestion with:

  • Header: "Create Task"
  • Question: "Describe the task, bug, or idea you want to capture. You can paste a client message, a rough note, or a voice transcription — I'll handle the formatting."

Step 2 — Classify the Task Type

If --type was explicitly provided, use that type and skip classification.

Otherwise, analyze the raw input and classify it into one of the 5 types using these rules:

TypeWhen to use
[US]Describes a new feature or capability from a user's perspective. Focus is on what the user wants to achieve.
[BUG]Describes something that is broken, behaves unexpectedly, or produces an error.
[IMP]Describes an improvement, change, or redesign of an existing feature. Not a new feature and not a bug.
[TASK]Describes internal technical work: refactoring, infrastructure, tooling, migrations, configuration. No user-facing story.
[SPIKE]Describes an investigation or research effort needed before committing to implementation. Centers on a specific question.

If the input clearly maps to one type, proceed silently.

If the input is ambiguous between two types (e.g., IMP vs US), state your classification and reasoning in one line before continuing:

"I'm classifying this as [IMP] because it modifies an existing feature rather than introducing new functionality."


Step 3 — Read the Corresponding Template

Based on the classified type, read the template file:

TypeTemplate file
[US]templates/clickup/us_task_template.md
[BUG]templates/clickup/bug_task_template.md
[IMP]templates/clickup/imp_task_template.md
[TASK]templates/clickup/task_task_template.md
[SPIKE]templates/clickup/spike_task_template.md

Use the Read tool to load the template. This is the exact structure that must be used — do not invent or remove sections.


Step 4 — Fill Out the Template

Populate every field in the template using only information present in the input. Apply the master prompt rules strictly:

  • Required fields: must be filled. If the input does not provide enough information, mark the field with ⚠ MISSING: [description of what's missing] — never leave a required field blank or invent content.
  • Optional fields: fill them only if the input contains relevant information. If there is nothing to say, omit the entire optional section from the output.
  • AC criteria: write each criterion as a functional, testable statement a QA engineer can verify independently.
  • "Summary": always include this section at the end. Summarize in 3 lines max: what to do, why it matters, and how to confirm it's done.
  • Open questions: if the input is ambiguous about scope, add an "Open Questions" section at the end listing each uncertainty as a numbered question.

Produce the fully filled-out ticket as a markdown block.


Step 5 — Confirm with User

Present the filled-out ticket to the user and ask:

"Here is the ticket I'm going to create in ClickUp. Does this look correct? Let me know if you'd like any changes before I create it."

Incorporate any requested changes. Repeat if necessary until the user confirms.


Step 6 — Select Target List in ClickUp

Fetch the workspace hierarchy to identify the correct list:

mcp__clickup__clickup_get_workspace_hierarchy

Present the available spaces and lists in a readable format. Use AskUserQuestion to ask:

"Which ClickUp list should I create this ticket in?"

Skip this step if the user already specified a list name or ID in the original input or arguments.


Step 7 — Create the Task in ClickUp

Create the task using:

mcp__clickup__clickup_create_task {
  list_id: "<selected list id>",
  name: "<ticket title>",
  description: "<full ticket body in markdown — all sections>"
}

The description must contain the complete filled-out template — all sections in order, exactly as confirmed in Step 5. Do not truncate, summarize, or omit any section.

Extract the url from the response.


Step 8 — Report

Output the result in this exact format:

✅ Task successfully created in ClickUp.

Type:   <[US] | [BUG] | [IMP] | [TASK] | [SPIKE]>
Title:  <task title>
URL:    <clickup task url>

If any required fields were marked with ⚠ MISSING:, include a reminder:

⚠ Incomplete fields: some required fields could not be filled with the available information.
  Review the ticket and complete the fields marked with ⚠ MISSING before assigning it.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.