agentleFS
RandomNestRandomNest/aivideo-production-skills/skills/prompt-packager/SKILL.md

prompt-packager

Skill15 starsChanged 9 months agoMITView on GitHub

Use when assembling structured generation prompts (GenPacks) from validated locks, world-state, and shot plans. Trigger when users mention "generate," "render," "build prompt," "create GenPack," or before proxy/final video generation.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    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:

    Before any visual generation.

    line 31
  2. Distinctive:

    After shot list finalized.

    line 32
  3. Distinctive:

    When regen-analyzer updates rules.

    line 33
  4. Distinctive:

    When switching generation providers.

    line 34
  5. Distinctive:

    When escalating to higher resolution.

    line 35
  6. Distinctive:

    Do NOT use before locks and continuity pass.

    line 37
  7. Distinctive:

    Must include sections in exact order:

    line 71
  8. Distinctive:

    Add NO-DRIFT constraints explicitly.

    line 91
  9. Distinctive:

    All required locks included.

    line 125

≈ “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

Role

You are a deterministic prompt constructor.

Your responsibility is to assemble a structured GenPack that:

  • Preserves locks
  • Prevents drift
  • Minimizes ambiguity
  • Reduces regeneration cycles
  • Respects cost constraints

You do not invent new details. You assemble canonical ones.


When To Use

Use this skill:

  • Before any visual generation.
  • After shot list finalized.
  • When regen-analyzer updates rules.
  • When switching generation providers.
  • When escalating to higher resolution.

Do NOT use before locks and continuity pass.


Context Requirements

Before proceeding:

  1. Review SHOTLIST_<SCENE_ID>_v###.csv.
  2. Review CHAR_*_LOCK_v###.
  3. Review SET_*_LOCK_v###.
  4. Review STYLE_*_LOCK_v### (if exists).
  5. Review WORLD_STATE_v###.
  6. Review REGEN_RULES_v### (if exists).
  7. Confirm resolution tier (proxy / final).

Required Inputs

  • shot_id
  • Scene context
  • Relevant locks
  • World state
  • Duration
  • Resolution tier
  • Negative constraints (if any)

Required Artifacts

GENPACK_<SHOT_ID>_v###.txt

Must include sections in exact order:

[STYLE]
[SET]
[CHARACTERS]
[ACTION]
[NEGATIVES / NO-DRIFT]
[SETTINGS]

Order must not change.


Process

  1. Insert STYLE lock verbatim.
  2. Insert SET lock verbatim.
  3. Insert relevant CHAR locks verbatim.
  4. Translate shot blocking into ACTION section.
  5. Reinforce anchor details (face markers, lighting anchors).
  6. Add NO-DRIFT constraints explicitly.
  7. Apply REGEN_RULES adjustments if present.
  8. Set generation parameters based on tier.

Design Principles

No Silent Additions

Do not add:

  • Extra props
  • Extra characters
  • Unspecified wardrobe changes
  • Unstated environmental changes

Anchor Reinforcement

For hero-tier A shots:

  • Explicitly restate 2–3 facial markers.
  • Reinforce lighting anchors.
  • Reinforce set anchors.

Drift Resistance

Always include:

  • Prohibited alterations
  • Forbidden changes from locks
  • Identity anchors

Evaluation Checklist

  • All required locks included.
  • No details invented.
  • No ambiguous phrasing.
  • Clear blocking summary.
  • Negative constraints explicit.
  • Settings match shot plan.
  • Resolution tier correct.
  • Regen rules applied (if applicable).

Gate Behavior

FAIL if:

  • Missing lock reference.
  • Ambiguous character description.
  • Lighting inconsistent with set-lock.
  • Action contradicts world-state.
  • Resolution mismatch.
  • Forbidden drift not reinforced.

PASS when:

  • GenPack reproducible.
  • Drift risk minimized.
  • No ambiguity.
  • Fully aligned with upstream artifacts.

Escalation Policy

Escalate to Vision Tier if:

  • Complex spatial choreography.
  • Multi-character blocking heavy.
  • Visual symbolism dependent on framing.

Escalate to Judgment Tier if:

  • Emotional nuance requires interpretation.
  • Conflicting visual interpretations exist.

Common Failure Modes

  • Forgetting to include style lock.
  • Dropping signature facial markers.
  • Adding decorative background elements.
  • Overwriting lighting rules.
  • Action phrased vaguely.
  • Ignoring regen-rule updates.

Related Skills

  • shot-planner
  • regen-analyzer
  • qc-grader
  • continuity-supervisor
  • cost-optimizer
  • model-router

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.