agentleFS
sb-devsb-dev/narrative-production-skills/skills/narrative-pack-create/SKILL.md

narrative-pack-create

Skill2 starsChanged 3 months agoMITView on GitHub

Create or revise self-contained Narrative Production customisation/extension packs. Use when defining a reusable medium, genre, style, audience, voice-cast, evaluation, or cross-project production profile that should be packaged as an Agent Skill rather than embedded in one project prompt.

The file

Show its rules and text

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈16

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create coherent Narrative Production customisation packs from proven production needs.

    line 8
  2. Distinctive:

    Create a self-contained Agent Skill package with only the references/assets it needs.

    line 19
  3. Distinctive:

    Create at least one showcase example whose README.md contains the exact generation prompt.

    line 20
  4. Distinctive:

    Create behavioural evals covering activation, application, refinement, precedence, and boundary behaviour.

    line 21
  5. Distinctive:

    Prefer a coherent ready-made pack over one skill per medium, genre, style, or voice dimension.

    line 26
  6. Distinctive:

    Do not create a new pack when an existing catalogue pack plus project-specific instructions is sufficient.

    line 27
  7. Distinctive:

    Never silently reopen or rewrite approved narrative decisions to satisfy a pack.

    line 30
  8. Distinctive:

    Do not use an individual creator's name as the whole style definition or attempt to reproduce an existing work exactly.

    line 32
  9. Distinctive:

    Keep voice casting optional unless spoken performance is intrinsic to the medium.

    line 33
  10. Distinctive:

    Never embed API keys, private voice assets, or unlicensed voice references.

    line 34
  11. Distinctive:

    Do not invent provider voice IDs.

    line 35
  12. Distinctive:

    Do not duplicate provider execution logic.

    line 36
  13. Distinctive:

    Every installable pack must be self-contained.

    line 37
  14. Distinctive:

    Every pack must include a showcase generation prompt and behavioural evals.

    line 38
  15. Distinctive:

    Add only files the pack genuinely needs.

    line 54
  16. Distinctive:

    Read the local references and templates when authoring or reviewing a pack.

    line 104

≈ “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

Narrative Pack Create

Create coherent Narrative Production customisation packs from proven production needs.

Operating sequence

  1. Identify the production need and inspect the existing extension-pack catalogue before proposing a new pack.
  2. Decide whether the need justifies a reusable pack or belongs in project-specific instructions.
  3. Define one coherent production profile: medium, genre, operational style, audience when relevant, and optional voice casting.
  4. Separate hard medium constraints from softer genre/style defaults.
  5. Define how the pack influences narrative-develop, narrative-write, narrative-continuity, narrative-evaluate, and narrative-revise only where relevant.
  6. Define downstream handoffs without taking over Comic, Video, Music/Audio, Game, or Web implementation responsibilities.
  7. Define pack-aware evaluation: intentional traits to preserve, defects to reject, and conditions that should not be penalised.
  8. Create a self-contained Agent Skill package with only the references/assets it needs.
  9. Create at least one showcase example whose README.md contains the exact generation prompt.
  10. Create behavioural evals covering activation, application, refinement, precedence, and boundary behaviour.
  11. Validate the package against the acceptance rules before treating it as ready.

Core rules

  • Prefer a coherent ready-made pack over one skill per medium, genre, style, or voice dimension.
  • Do not create a new pack when an existing catalogue pack plus project-specific instructions is sufficient.
  • Medium constraints may be hard; genre and style guidance are normally soft.
  • Explicit project instructions and approved artifacts override pack defaults.
  • Never silently reopen or rewrite approved narrative decisions to satisfy a pack.
  • A style must be operational: describe observable production characteristics, not merely a label.
  • Do not use an individual creator's name as the whole style definition or attempt to reproduce an existing work exactly.
  • Keep voice casting optional unless spoken performance is intrinsic to the medium.
  • Never embed API keys, private voice assets, or unlicensed voice references.
  • Do not invent provider voice IDs. Use placeholders or casting direction when a licensed/configured voice is unavailable.
  • Do not duplicate provider execution logic.
  • Every installable pack must be self-contained; repository-level docs are not runtime dependencies.
  • Every pack must include a showcase generation prompt and behavioural evals.

Minimum output

skills/<pack-name>/
├── SKILL.md
├── references/
│   └── production-profile.md
└── evals/
    └── evals.json

examples/extension-packs/<pack-name>/
└── README.md

Add only files the pack genuinely needs.

Pack profile

A pack should define:

identity
intended use
medium
genre
operational style
audience when relevant
optional voice casting
hard constraints
defaults
artifact/workflow effects
continuity effects
evaluation behaviour
revision behaviour
downstream handoffs
external requirements
conflicts / incompatibilities

Precedence

1. explicit project instructions
2. approved narrative artifacts and approved production decisions
3. selected extension pack
4. Narrative Production defaults

Surface conflicts rather than silently resolving them against stronger decisions.

Evaluation

A created pack is not ready merely because its files exist.

Verify that it:

  • activates when requested and stays inactive otherwise;
  • changes relevant production behaviour rather than only adding labels;
  • preserves approved work;
  • adapts evaluation to intentional characteristics without hiding real defects;
  • respects medium and project boundaries;
  • handles voice casting safely where present;
  • can be demonstrated by its showcase prompt.

Read the local references and templates when authoring or reviewing a pack.

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.