agentleFS
BitslixBitslix/BLXCode/.agents/skills/manage-plans/SKILL.md

manage-plans

Skill15 starsChanged 5 months agoMITView on GitHub

Create, update, and maintain BLXCode workspace plans under `.agents/plans/`. Use when Codex needs to add a new implementation plan, refine an existing plan, update plan task status, keep `PLANS.md` indexed, or explain the repository's Markdown plan format and task syntax to agents.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 12 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

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

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

    Keep .agents/plans/PLANS.md as the protected index.

    line 11
  2. Distinctive:

    Use one focused plan folder per feature, refactor, investigation, or multi-step implementation.

    line 12
  3. Very specific:

    Prefer lowercase hyphenated folder slugs with a canonical plan.md, for example kanban-board-view/plan.md.

    line 13
  4. Distinctive:

    Keep plans agent-actionable: concrete enough to implement, but not bloated with unrelated analysis.

    line 14
  5. Distinctive:

    Update the plan as work changes.

    line 15
  6. Distinctive:

    Use this default structure for new plans:

    line 19
  7. Distinctive:

    Use exactly one task per line:

    line 54
  8. Very specific:

    Check .agents/plans/PLANS.md and existing plan slugs to avoid duplicates.

    line 76
  9. Very specific:

    Create a focused plan.md file under .agents/plans/<slug>/.

    line 77
  10. Distinctive:

    Add a clear H1 title and the standard sections above.

    line 78
  11. Very specific:

    Add concrete ## Tasks entries using stable task IDs.

    line 79
  12. Very specific:

    Add a row to .agents/plans/PLANS.md with status, link, and concise description.

    line 80
  13. Very specific:

    Use planned, active, blocked, done, or cancelled in the index status column.

    line 88
  14. Distinctive:

    Update task status markers when work progresses.

    line 92
  15. Distinctive:

    Add tasks when the implementation discovers real new work.

    line 93
  16. Very specific:

    Remove tasks only when they are invalid or superseded.

    line 94
  17. Distinctive:

    Keep the index row status aligned with the plan's actual state.

    line 95
  18. Distinctive:

    Avoid reformatting the whole file unless explicitly asked.

    line 97
  19. Distinctive:

    Avoid vague tasks like "fix stuff" or "improve UI".

    line 102
  20. Distinctive:

    Do not include implementation details that are merely guesses if the repo can answer them.

    line 106

≈ “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.

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

Manage BLXCode Plans

Core Rules

  • Store durable plans as .agents/plans/<slug>/plan.md.
  • Keep .agents/plans/PLANS.md as the protected index. Do not delete or rename it.
  • Use one focused plan folder per feature, refactor, investigation, or multi-step implementation.
  • Prefer lowercase hyphenated folder slugs with a canonical plan.md, for example kanban-board-view/plan.md.
  • Keep plans agent-actionable: concrete enough to implement, but not bloated with unrelated analysis.
  • Update the plan as work changes. A stale plan is worse than no plan.

Plan File Structure

Use this default structure for new plans:

# Short Plan Title

## Summary

Briefly state the goal, current direction, and important constraints.

## Decisions

- Record product or technical decisions that should not be rediscovered.
- Include assumptions when the user has not explicitly decided something.

## Implementation Notes

- Describe the intended approach by subsystem or behavior.
- Mention important files only when needed to avoid ambiguity.
- Include public APIs, storage formats, commands, or UI contracts when they change.

## Tests

- List backend, frontend, integration, and manual checks expected for the work.

## Tasks

- [ ] `stable-task-id` - Concrete task title

Small plans may omit Decisions or Implementation Notes if they would be empty. Keep Summary, Tests, and Tasks whenever the plan is meant to guide implementation.

Task Syntax

BLXCode parses plan tasks from a ## Tasks section. ## Todos is accepted, but prefer ## Tasks.

Use exactly one task per line:

- [ ] `task-id` - Pending task
- [>] `active-task` - In-progress task
- [!] `blocked-task` - Blocked task
- [x] `done-task` - Completed task
- [-] `cancelled-task` - Cancelled task

Status markers:

  • [ ] pending
  • [>] in progress
  • [!] blocked
  • [x] completed
  • [-] cancelled

Task IDs must be stable, unique within the plan, lowercase where practical, and wrapped in backticks. Do not rename a task ID casually because BLXCode uses it to sync plan-linked tasks.

Creating A New Plan

  1. Check .agents/plans/PLANS.md and existing plan slugs to avoid duplicates.
  2. Create a focused plan.md file under .agents/plans/<slug>/.
  3. Add a clear H1 title and the standard sections above.
  4. Add concrete ## Tasks entries using stable task IDs.
  5. Add a row to .agents/plans/PLANS.md with status, link, and concise description.

Index row format:

| planned | [new-plan/plan.md](new-plan/plan.md) | One-sentence description of the intended work |

Use planned, active, blocked, done, or cancelled in the index status column. Match the style already used in PLANS.md.

Updating Existing Plans

  • Update task status markers when work progresses.
  • Add tasks when the implementation discovers real new work.
  • Remove tasks only when they are invalid or superseded; prefer [-] for cancelled work with useful history.
  • Keep the index row status aligned with the plan's actual state.
  • Preserve unrelated sections and user-authored content.
  • Avoid reformatting the whole file unless explicitly asked.

Quality Bar

  • Make each task independently actionable.
  • Avoid vague tasks like "fix stuff" or "improve UI".
  • Include edge cases and acceptance checks in Tests, not as hidden assumptions.
  • For UI work, include expected user-visible behavior.
  • For storage/API work, include compatibility and migration expectations.
  • Do not include implementation details that are merely guesses if the repo can answer them; inspect first.

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.