pickle-ref

pickle-ref is a command for Claude Code from MrJPTech/macro-pickle. It costs 23 tokens per session (751 once invoked), scanned A, original, MIT.

A command for cleaning and regenerating a product's reference images, including studio and lifestyle views, while keeping the original files backed up.

In plain words
What is it for?
Use it to create consistent product angles, add real-world objects that show scale, and prepare cleaner references for image or video generation.
Why use it?
It addresses messy or poorly scaled reference images that can make a product look the wrong size in generated video scenes.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/mrjptech/macro-pickle/pickle-ref
Clone the repo
git clone --depth 1 https://github.com/MrJPTech/macro-pickle

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for pickle-ref

README.md
[![agentmods](https://agentmods.dev/badge/commands/mrjptech/macro-pickle/pickle-ref.svg)](https://agentmods.dev/commands/mrjptech/macro-pickle/pickle-ref)
Your own site
<a href="https://agentmods.dev/commands/mrjptech/macro-pickle/pickle-ref"><img src="https://agentmods.dev/badge/commands/mrjptech/macro-pickle/pickle-ref.svg" alt="Measured on agentmods" height="20"></a>
Per session 23 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 751 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00023 $0.00751
Opus 5 $0.00012 $0.00376
Sonnet 5 $0.00005 $0.00150
Haiku 4.5 $0.00002 $0.00075

Measured 3d ago against content hash a1f74642ba34, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

pickle-ref scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/pickle-ref.md · 49 lines

How it starts

The opening of the file, as written. The whole thing — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/pickle-ref

Regenerate a product's reference frames when the studio shot is messy, off-brand, or — the most common case — causes Veo to OVERSIZE the product in clips. Isolated white-background studio refs carry no scale cue, so a palm-sized tin renders bucket-sized in a scene. This is the reference-level fix; it pairs with the per-SKU _scale anchor in the clip scenes config.

Workflow (proven on the mosquito balm jar, 2026-06-30)

  1. Look at the current refsstudio_01..03 + life_01..02 in the gen5 dir (extract a frame / open the PNGs; downscale >2000px with ffmpeg before viewing).
  2. Regenerate clean STUDIO refs via nano, feeding the OLD refs to keep identity:
    pnpm nano --prompt "<clean SINGLE product, explicit true scale + form factor, pure white seamless, soft studio light, no text/logos>" \
      --refs "<gen5>/studio_01.png,<gen5>/studio_02.png,<gen5>/studio_03.png" \
      --model 2 --aspect 1:1 --size 2K --count 2 --out "<review-dir>"
    
    Generate 2–3 consistent angles for a clean ASSET set. Review every one before installing.
  3. Regenerate LIFESTYLE refs the same way, but in-scene with explicit scale objects (pinecones, a coin, pine needles, normal-sized props) — lifestyle frames are where scale context belongs (you can't put scale objects in the clean ASSET studio refs).
  4. Back up originals, then install:
    mkdir -p "<gen5>/_orig-<desc>"
    cp "<gen5>"/studio_0*.png "<gen5>"/life_0*.png "<gen5>/_orig-<desc>/"   # backup
    cp "<review>/<keeper>.png" "<gen5>/studio_01.png"   # …install the keepers
    
    Never overwrite gen5 originals without a backup.
  5. Re-render the clip/pickle-clips --skus <sku> once Veo quota is open (Flow now, or API on reset), and add a _scale entry for the SKU in content/clip-scenes/<store>.json.

Why white-bg refs oversize

Veo's ASSET lock preserves appearance, not real-world scale. Fix at three layers: (a) a clean ref with an unambiguous form factor, (b) the per-SKU _scale phrase in the clip scenes config, (c) in-frame scale objects in the lifestyle frames.

Read the full file on GitHub · 49 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 49 lines · 23 tokens per session scan A a1f74642ba34

Subscribe to this mod's changes

pickle-ref is a command published in the GitHub repository MrJPTech/macro-pickle (2 stars, last pushed 1mo ago), licensed MIT. It adds 23 tokens to every session and 751 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.