agentleFS
ShiningPlagueShiningPlague/shiningplague-studio/skills/help/SKILL.md

help

Skill0 starsChanged 4 months agoMITView on GitHub

Use when designer asks 'where are we', 'what should I do now', 'what's next', 'I'm stuck', 'status', OR at session start to surface canonical state. Analyzes what is done and the user's query and offers advice on what to do next. ShiningPlague-adopted: adds workstream + adoption-TODO awareness + paused-spec surfacing on top of canonical phase logic.

76 other projects use a version of this, 9 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use soak-test, team-live-ops and 2 more.

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.
  • lilangMaxSkill

    soak-test

    Generate a soak test protocol for extended play sessions. Defines what to observe, measure, and log during long play sessions to surface slow leaks, fatigue effects, and edge cases that only appear after sustained play. Primarily used in Polish and Release phases.

    Also in 7 projects that use this

  • lilangMaxSkill

    team-live-ops

    Orchestrate the live-ops team for post-launch content planning: coordinates live-ops-designer, economy-designer, analytics-engineer, community-manager, writer, and narrative-director to design and plan a season, event, or live content update.

    Also in 7 projects that use this

  • +2 more from lilangMax

The file

Show its rules and text

We split this skill into the rules it gives an agent

17 rules. 0 are common, 8 are distinctive, and 9 are very specific to this project.

  • 0

    Common

    Most of the 57 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 9

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After a major skill fires — recommend the next canonical step

    line 38
  2. Very specific:

    Read .claude/docs/workflow-catalog.yaml.

    line 44
  3. Very specific:

    Read production/session-state/active.md.

    line 111
  4. Very specific:

    If step has artifact.glob:

    line 129
  5. Distinctive:

    Use Glob to check

    line 130
  6. Very specific:

    If min_count specified, verify count

    line 131
  7. Very specific:

    If artifact.pattern specified, use Grep to verify pattern

    line 132
  8. Very specific:

    If step has artifact.note (no glob):

    line 136
  9. Very specific:

    If step has no artifact field:

    line 139
  10. Very specific:

    When current phase is production, check production/sprint-status.yaml before glob-based story checks.

    line 144
  11. Distinctive:

    Keep it short and direct. Quick orientation, not a report.

    line 171
  12. Distinctive:

    Keep the tone friendly and orienting — this is a helpful signpost, not a compliance report.

    line 215
  13. Distinctive:

    If all required steps in current phase are complete (or nearly), add:

    line 221
  14. Distinctive:

    If multiple required steps remain, skip the gate warning.

    line 224
  15. Distinctive:

    If user seems stuck or confused, add:

    line 230
  16. Distinctive:

    Only show if input suggested confusion ("I don't know", "stuck", "lost", "not sure").

    line 240
  17. Very specific:

    After firing this skill, if user picks canonical recommendation, the natural next skill fires per team-orchestrator.md activity chain.

    line 246

“Common” is measured: kept by at least 6 in 10 of the 57 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

Studio Help — What Do I Do Next?

🌱 ShiningPlague-adopted. Originally Donchitos; battle-tested on a shipped Godot project. Upstream canonical phase logic preserved + workstream/adoption awareness layered on.

This skill is read-only — it reports findings but writes no files.

Figures out where you are in the game development pipeline and tells you what comes next. Lightweight — not a full audit. For full gap analysis, use /project-stage-detect.

If an artifact named here is absent: say so plainly in one line, skip that step, and continue. Never invent the file to satisfy a checklist, and never fail a close because an optional artifact was never created.

When to fire

  • Designer asks: "where are we", "what should I do now", "what's next", "I'm stuck", "I don't know", "status", "help"
  • Session-start hook output PHASE ARTIFACT STATUS surfaced an unresolved gap
  • After a major skill fires — recommend the next canonical step

Step 1: Read the Catalog

Read .claude/docs/workflow-catalog.yaml. Authoritative list of all phases, their steps, required/optional flags, and artifact globs.

Step 1b: Find Skills Not in the Catalog

Glob .claude/skills/*/SKILL.md. For each, extract the name: field from frontmatter.

Compare against command: values in the catalog. Any skill whose name doesn't appear in the catalog is uncataloged — usable but not phase-gated.

Collect for Step 7 footer:

### Also installed (not in workflow)
- `/skill-name` — [description from SKILL.md frontmatter]

Show only if at least one uncataloged skill exists. Limit to the 10 most relevant for current phase.


Step 2: Determine Current Phase

  1. Read production/stage.txt — authoritative. Map to catalog phase key:
    • "not-started" → the seeded default, not a phase. No gate has been cleared; fall through to step 2 and infer from artifacts.
    • "Concept" → concept
    • "Systems Design" → systems-design
    • "Technical Setup" → technical-setup
    • "Pre-Production" → pre-production
    • "Production" → production
    • "Polish" → polish
    • "Release" → release
  2. If missing (a --no-scaffold install) or still not-started, infer phase from artifacts (most-advanced match wins). The installer seeds this stack blank, so judge these on content, not existence — an unfilled template is not an artifact, and a file still carrying the scaffold-seed: unwritten marker near the top counts as absent (.claude/docs/doc-stack.md § Seeded is not written):
    • src/ has 10+ source files → production
    • production/epics/**/*.md exists → pre-production
    • docs/adr/[0-9]*.md exists → technical-setup (NNN-<slug>.md only; the leading digit excludes the seeded docs/adr/TEMPLATE.md)
    • docs/gdd/systems-index.md has rows in its systems table → systems-design
    • docs/gdd/game-concept.md has a written elevator pitch (not the bracketed […] prompt text) → concept
    • Nothing → concept (fresh project)

Step 3: Read Session Context

Read production/session-state/active.md. Extract:

  • What was most recently worked on
  • In-progress tasks or open questions
  • Current epic/feature/task from STATUS block

This personalizes the output.


Step 4: Read Per-Workstream State (studio addition)

Glob production/workstreams/*.md. For each, extract:

  • Current Phase: line
  • In Progress section content
  • Blocked section content
  • Open Questions section content

Report per-workstream phase + what's in flight + blockers.


Step 5: Read Open Adoption TODOs (studio addition)

Read production/session-state/active.md. Look for 🚨 OPEN ADOPTION PLAN TODOs section. List unresolved items.

Also check docs/adoption-plan-*.md for any unchecked items not yet surfaced.


Step 6: Read Open Specs (studio addition)

Glob docs/specs/*.md. For each, extract Status (🚧 IN PROGRESS / ✅ FINAL / 🗄️ ARCHIVED). List in-progress specs — paused work waiting to resume.


Step 7: Check Step Completion for Current Phase

For each step in current phase (from catalog):

Artifact-based checks

If step has artifact.glob:

  • Use Glob to check
  • If min_count specified, verify count
  • If artifact.pattern specified, use Grep to verify pattern
  • Complete = artifact condition met
  • Incomplete = artifact missing or pattern not found

If step has artifact.note (no glob):

  • Mark MANUAL — ask user

If step has no artifact field:

  • Mark UNKNOWN — not trackable

Special case: production phase — read sprint-status.yaml

When current phase is production, check production/sprint-status.yaml before glob-based story checks. The installer seeds it status: none with an empty stories: [] — that is "no sprint has been planned yet", not a sprint with no stories. Once a sprint exists, read it:

  • Stories status: in-progress → "currently active"
  • status: ready-for-dev → "next up"
  • status: done → count as complete
  • status: blocked → surface as blocker

Skip glob check for implement and story-done steps — YAML is authoritative.

Special case: repeatable: true (non-production)

For repeatable steps outside production (e.g. "System GDDs"), artifact check tells whether any work has been done. Label differently — show detected, note it may be ongoing.


Step 8: Determine Recommended Next Canonical Action

Priority order:

  1. Project canonical phase gap (foundational) — if Concept/Systems-Design artifacts missing, recommend canonical skill (e.g., /design-system game-concept if pointer-only).
  2. Workstream-specific work — if foundational gaps closed, check which workstream the designer was last on (active.md priorities[0]) and recommend the next canonical step.
  3. In-progress paused work — if a spec is 🚧 IN PROGRESS, recommend resuming.
  4. Open adoption TODOs — surface unresolved items even if recommending phase work.

Step 9: Present Output

Keep it short and direct. Quick orientation, not a report.

## Where You Are: [Phase Label]

**In progress:** [from active.md, if any]

### ✓ Done
- [completed step name]

### → Next up (recommended)
**[Step name]** — [description]
Command: `[/command]`

### Workstream Status

| Workstream | Phase | In Progress | Blocked |
|---|---|---|---|
| Combat | Production | [item] | — |
| Narrative | Concept | — | [blocker] |

### Open Adoption TODOs
- [item 1]
- [item 2]

### Open Specs (paused)
- 🚧 [spec path] — [topic]

### ~ Also available (optional)
- **[Step name]** — [description] → `/command`

### Coming up after that
- [Next step name] (`/command`)

---
Getting close to the **[next phase]** gate → run `/gate-check` when you're ready.

Formatting rules:

  • ✓ for confirmed complete
  • → for the recommended next step (only one — the first blocker)
  • ~ for optional steps available now
  • Show commands inline as backtick code
  • For MANUAL steps, ask: "I can't tell if [step] is done — has it been completed?"
  • Keep the tone friendly and orienting — this is a helpful signpost, not a compliance report.

Step 10: Gate Warning (if close)

If all required steps in current phase are complete (or nearly), add:

"You're close to the [Current] → [Next] gate. Run /gate-check when ready."

If multiple required steps remain, skip the gate warning.


Step 11: Escalation Paths

If user seems stuck or confused, add:

---
Need more detail?
- `/project-stage-detect` — full gap analysis with all missing artifacts listed
- `/gate-check` — formal readiness check for your next phase
- `/start` — re-orient from scratch

Only show if input suggested confusion ("I don't know", "stuck", "lost", "not sure").


Cross-skill awareness

  • After firing this skill, if user picks canonical recommendation, the natural next skill fires per team-orchestrator.md activity chain.
  • Pair with /project-stage-detect (also project-local) — that skill does the deep gap audit; this one does the day-to-day "what's next" framing.

Reference files (project)

  • .claude/docs/workflow-catalog.yaml — canonical phase pipeline
  • production/stage.txt — current canonical phase
  • production/workstreams/*.md — per-workstream state
  • production/session-state/active.md — open TODOs + priorities
  • docs/specs/*.md — paused brainstorm specs
  • docs/adoption-plan-*.md — historical adoption plans written by /adopt
  • tools/generate_session_context.sh — OPTIONAL, project-owned. The template never ships it; .claude/hooks/session-start.sh runs it only if the project wrote one. Absent by default.

Failure modes

  • ❌ Listing all phase artifacts as plain status — that's /project-stage-detect's job. This skill RECOMMENDS next action.
  • ❌ Asking "what do you want to do?" without recommending. Always recommend.
  • ❌ Forgetting to read active.md adoption TODOs section.
  • ❌ Re-scanning all globs when session-start hook already did it.

Collaborative Protocol

  • Never auto-run the next skill. Recommend it, let the user invoke it.
  • Ask about MANUAL steps rather than assuming complete or incomplete.
  • Match the user's tone — if they sound stressed ("I'm totally lost"), be reassuring and give one action, not a list of six.
  • One primary recommendation — user should leave knowing exactly one thing to do next.

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.