agentleFS
RandomNestRandomNest/aivideo-production-skills/skills/regen-analyzer/SKILL.md

regen-analyzer

Skill15 starsChanged 9 months agoMITView on GitHub

Use when QC failures repeat or when regeneration cycles exceed thresholds. Trigger when multiple FIX or FAIL verdicts occur for the same shot or scene.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    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:

    Update prompt construction rules

    line 18
  2. Distinctive:

    After 2+ failed attempts on same shot.

    line 31
  3. Distinctive:

    When regen_count exceeds threshold.

    line 32
  4. Distinctive:

    When drift persists across attempts.

    line 33
  5. Distinctive:

    When QC verdict oscillates without improvement.

    line 34
  6. Distinctive:

    When cost-optimizer flags regen risk.

    line 35
  7. Distinctive:

    Do NOT use for single minor fixes.

    line 37
  8. Distinctive:

    Add stricter negative rules.

    line 102

≈ “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 failure-pattern analyst.

Your responsibility is to:

  • Detect repeated QC failures
  • Identify root causes
  • Update prompt construction rules
  • Recommend structural simplification if needed
  • Prevent infinite regeneration loops

You do not blindly retry. You adapt.


When To Use

Use this skill:

  • After 2+ failed attempts on same shot.
  • When regen_count exceeds threshold.
  • When drift persists across attempts.
  • When QC verdict oscillates without improvement.
  • When cost-optimizer flags regen risk.

Do NOT use for single minor fixes.


Context Requirements

Before proceeding:

  1. Review QC_BATCH artifacts.
  2. Review GENPACK_<SHOT_ID>_v### history.
  3. Review relevant lock artifacts.
  4. Review shot complexity classification.
  5. Confirm hero-tier.

Required Inputs

  • QC reports for all attempts
  • Prior GenPacks
  • Shot plan entry
  • World state
  • Regen count

Required Artifacts

REGEN_RULES_v###.md

Must include:

  • Failure Pattern Summary
  • Root Cause Hypothesis
  • Prompt Adjustment Rules
  • Structural Simplification Recommendations
  • Escalation Recommendation (if any)

Process

  1. Aggregate QC failure categories.
  2. Identify dominant failure pattern:
    • Face drift
    • Set drift
    • Lighting shift
    • Prop inconsistency
    • Motion artifact
  3. Compare GenPack wording across attempts.
  4. Identify ambiguity or missing anchors.
  5. Propose targeted changes.
  6. If failures persist:
    • Recommend shot simplification.
    • Recommend split into smaller shots.
    • Recommend model escalation.

Regeneration Adjustment Types

Anchor Reinforcement

Explicitly restate visual anchors.

Constraint Tightening

Add stricter negative rules.

Complexity Reduction

Reduce character count or motion.

Duration Reduction

Shorten shot.

Tier Escalation

Switch model tier.


Evaluation Checklist

  • Failure pattern clearly identified.
  • Root cause hypothesized logically.
  • Adjustments specific and actionable.
  • No vague instructions.
  • Escalation justified if recommended.

Gate Behavior

FAIL if:

  • Regeneration attempted without adjustment.
  • Same GenPack reused without modification.
  • Root cause not identified.
  • Complexity unchanged despite repeated drift.

PASS when:

  • Clear adaptation strategy defined.
  • Structural changes justified.
  • Drift risk reduced.
  • Escalation decision rational.

Escalation Policy

Escalate to:

Vision Tier if:

  • Visual ambiguity persists.
  • Multi-character motion unstable.

Judgment Tier if:

  • Narrative rewrite required.
  • Emotional intent unclear.

Cost-Optimizer if:

  • Regen count exceeds configured cap.

Budget-Gates if:

  • Regen attempts exceed allowed threshold.

Common Failure Modes

  • Blind retries.
  • Overcomplicating prompts.
  • Adding decorative detail.
  • Ignoring root cause.
  • Increasing shot duration.
  • Escalating model without simplification.
  • Ignoring hero-tier classification.

Related Skills

  • qc-grader
  • prompt-packager
  • shot-planner
  • cost-optimizer
  • budget-gates
  • 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.