agentleFS
wtsi-hgiwtsi-hgi/agentskills/skills/phase-creator/SKILL.md

phase-creator

Skill1 starsChanged 39 days agoMITView on GitHub

Creates phase plan documents from a spec.md Implementation Order. Invoked by spec-writer, not directly.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read and follow agent-conduct and unslop before starting.

    line 8
  2. Distinctive:

    Read spec's Implementation Order.

    line 22
  3. Very specific:

    Create phase<N>.md per phase in the output directory (format below).

    line 25
  4. Very specific:

    If a batch depends on a prior batch: ### Batch <B> (parallel, after batch <B-1> is reviewed).

    line 74
  5. Distinctive:

    After all parallel items, include:

    line 77
  6. Distinctive:

    NEVER invent items not in the spec's Implementation Order.

    line 102
  7. Very specific:

    ALWAYS include both - [ ] implemented and - [ ] reviewed checkboxes.

    line 103
  8. Distinctive:

    ALWAYS identify parallel items and group them into batches.

    line 104
  9. Distinctive:

    ALWAYS follow unslop § Mechanics, and keep phase files ASCII only.

    line 105

≈ “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

Phase Creator Skill

Read and follow agent-conduct and unslop before starting.

You create one phase file per phase from a spec's Implementation Order, formatted for use by the orchestrator skill.

Input

  • Spec path - path to spec.md.
  • Output directory - for phase files (normally same dir as spec).
  • Implementor/reviewer skill names - e.g. go-implementor/go-reviewer (determine from project context if not provided).

Procedure

  1. Read spec's Implementation Order. Note every phase, title, and story IDs.
  2. Analyse dependencies: group independent items into parallel batches. Dependent items go into later batches.
  3. Create phase<N>.md per phase in the output directory (format below).
  4. Return summary of files created and their items.

Phase File Format

# Phase <N>: <Phase title from spec>

Ref: [spec.md](spec.md) sections <comma-separated story IDs>

## Instructions

Use the `orchestrator` skill to complete this phase, coordinating
subagents with the `<implementor-skill>` and `<reviewer-skill>`
skills.

## Items

<items - see below>

Sequential items

### Item <N>.<M>: <Story ID> - <Story title>

spec.md section: <Story ID>

<Brief description referencing acceptance test count from spec.>

- [ ] implemented
- [ ] reviewed

Parallel batches

### Batch <B> (parallel)

#### Item <N>.<M>: <Story ID> - <Title> [parallel with <others>]

spec.md section: <Story ID>

<Brief description.>

- [ ] implemented
- [ ] reviewed

If a batch depends on a prior batch: ### Batch <B> (parallel, after batch <B-1> is reviewed).

After all parallel items, include:

For parallel batch items, use separate subagents per item under the
`subagents` skill's shared concurrency limits.
Launch review subagents using the `<reviewer-skill>` skill
(review all items in the batch together in a single review
pass).

Item descriptions

  • Name functions, types, or files to implement.
  • Reference spec.md section for full details.
  • State acceptance test count (e.g. "covering all 5 acceptance tests from A1").
  • Note dependencies on other items if relevant.

Numbering

  • Items: <phase>.<sequence> (e.g. 4.1, 4.2).
  • Batches: sequential within phase.
  • Sequence numbers continuous across batches.

Rules

  • NEVER invent items not in the spec's Implementation Order.
  • ALWAYS include both - [ ] implemented and - [ ] reviewed checkboxes.
  • ALWAYS identify parallel items and group them into batches.
  • ALWAYS follow unslop § Mechanics, and keep phase files ASCII only.

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.