agentleFS
HyunKNHyunKN/harness-kit/skills/planning-workflow/SKILL.md

planning-workflow

Skill0 starsChanged 7 months agoMITView on GitHub

Use when work is already clear enough to execute but still too large to implement safely without a written plan. Trigger for multi-step features, refactors, risky bug fixes, or coordinated agent work that should be broken into bounded tasks with validation checkpoints. Do not use to clarify vague goals; start with `spec-writing` when the request itself is still underspecified.

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the current issue or spec before changing code.

    line 14
  2. Distinctive:

    Prefer vertical slices over layer-by-layer work when possible.

    line 17
  3. Distinctive:

    Put acceptance and validation beside each task, not only at the end.

    line 18
  4. Very specific:

    Use the issue doc's Plan and Validation sections unless the plan is large enough to deserve its own note.

    line 19
  5. Distinctive:

    Stop once the next implementation steps are explicit enough to execute and track in the issue doc.

    line 20

“Common” is measured: kept by at least 6 in 10 of the 5 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

Planning Workflow

Break a spec or issue into small tasks that another Codex user can execute without re-scoping the problem.

This skill is for implementation sequencing, not initial problem discovery.

Default flow

  1. Read the current issue or spec before changing code.
  2. Identify dependencies, risky areas, and blockers.
  3. Split the work into small validated steps.
  4. Prefer vertical slices over layer-by-layer work when possible.
  5. Put acceptance and validation beside each task, not only at the end.
  6. Use the issue doc's Plan and Validation sections unless the plan is large enough to deserve its own note.
  7. Stop once the next implementation steps are explicit enough to execute and track in the issue doc.

Good fit

  • features that touch multiple files
  • work that will span more than one session
  • tasks that may involve sub-agents
  • fixes where implementation order matters

Avoid for

  • tiny edits with obvious scope
  • tasks where the issue doc already has a solid task list

Resources

  • See references/plan-checklist.md for a short planning rubric.
  • Pair with issue-driven-workflow to keep the plan attached to the real execution record.

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.