critic-loop

critic-loop is a skill for Claude Code from Zane-0x5a/remotion-director. It costs 119 tokens per session (852 once invoked), scanned A, original, MIT.

A selection-and-review stage for visual design work. It chooses the most promising image from several candidates, then repeats feedback and revisions between a critic and a builder until the result converges.

In plain words
What is it for?
Choosing a base from multiple rendered images and guiding repeated visual critique and implementation rounds in the remotion-director pipeline.
Why use it?
It provides a structured way to improve one chosen design instead of trying to refine every candidate. Keeping the critic involved across rounds helps preserve feedback between revisions.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the remotion-director plugin — 3 skills, 4 agents shipped together

Good fit Choosing a base from multiple rendered images and guiding repeated visual critique and implementation rounds in the remotion-director pipeline.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add Zane-0x5a/remotion-director
Claude Code
/plugin install remotion-director

Made for: Claude Code.

Or install remotion-director, the plugin that ships this one along with the rest of its 3 skills, 4 agents.

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 critic-loop

README.md
[![agentmods](https://agentmods.dev/badge/skills/zane-0x5a/remotion-director/critic-loop.svg)](https://agentmods.dev/skills/zane-0x5a/remotion-director/critic-loop)
Your own site
<a href="https://agentmods.dev/skills/zane-0x5a/remotion-director/critic-loop"><img src="https://agentmods.dev/badge/skills/zane-0x5a/remotion-director/critic-loop.svg" alt="Measured on agentmods" height="20"></a>
Per session 119 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 852 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00119 $0.00852
Opus 5 $0.00060 $0.00426
Sonnet 5 $0.00024 $0.00170
Haiku 4.5 $0.00012 $0.00085

Measured 8d ago against content hash 9c2ba85815eb, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

critic-loop 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 8d 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.

skills/critic-loop/SKILL.md · 31 lines

How it starts

The opening of the file, as written. The whole thing — 31 lines — stays where its author put it; the contents beside it link to each section on GitHub.

critic-loop

This stage turns N rendered draws into one refined piece. Two protocols govern it; both are the authoritative wording and are used verbatim (⟨…⟩ = slots the orchestrator fills, the rest unchanged). Do NOT paraphrase them — Read the files and use their exact text.

1 · Blind select — pick the most promising base

Read ${CLAUDE_PLUGIN_ROOT}/skills/critic-loop/BLIND-SELECT-PROTOCOL.md. Spawn the blind-selector agent (fresh, one-shot) with the brief + the N candidate dirs (each has 6 still-*.png + strip/). It selects for potential (the base whose ceiling after the loop is highest), not fewest current flaws — fixable execution nits must not count against a strong base. It returns { winner, reason }. The orchestrator does NOT judge; it hands over candidates and takes back the winner.

2 · Critic loop (甲乙环) — refine the winner until it converges

Read ${CLAUDE_PLUGIN_ROOT}/skills/critic-loop/CRITIC-PROTOCOL.md (条带规格 harness contract + the verbatim 甲 prompt + the per-round ferry message + 乙环纪律).

  • 甲 (critic) = the aesthetic-critic agent. Spawn ONE instance for the piece and continue the SAME instance across rounds (it is persistent — retained memory across rounds is the point — but design-blind: it sees only the frame strip + the brief, never DESIGN.md or code). It reports phenomena + severity and never prescribes a fix; its last line is CONVERGED: YES|NO.
  • 乙 (builder) = the SAME continuous-context builder instance that designed and built the winning draw (do NOT spawn a fresh agent to read-back context — that is the degraded rescue form). It receives 甲's verdict and adjudicates each item per §5 环纪律.
  • The orchestrator ferries verbatim, both ways, and never judges aesthetics:
    • 甲's verdict text → the builder's conversation, and archive it to ⟨RUN_DIR⟩/CRITIC-VERDICTS.md.
    • The builder's pixel-grounded rebuttal → 甲 (甲's INTEGRITY forbids it from reading any file other than the frames + its own crops, so it cannot fetch the rebuttal itself).
  • The builder re-renders each round to out/r⟨N⟩ (render-arm then render-strip), confirms non-white, appends fixes to ⟨RUN_DIR⟩/FIXES.md.
  • Loop until CONVERGED: YES — no round cap. 甲's retained cross-round memory is what drives fast convergence (typically a few rounds); never stop it while high-/med-severity items remain. The result is the latest out/rN.

Read the full file on GitHub · 31 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 8d ago First seen · 31 lines · 119 tokens per session scan A 9c2ba85815eb

Subscribe to this mod's changes

critic-loop is a skill published in the GitHub repository Zane-0x5a/remotion-director (28 stars, last pushed 16d ago), licensed MIT. It adds 119 tokens to every session and 852 once invoked, about $0.0006 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 2026-08-30.