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.
npx agentmods add agents/noelpuig/claude-explains/chapter-coordinatorgit clone --depth 1 https://github.com/noelpuig/claude-explainsWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00023 | $0.00907 |
| Opus 5 | $0.00012 | $0.00453 |
| Sonnet 5 | $0.00005 | $0.00181 |
| Haiku 4.5 | $0.00002 | $0.00091 |
Grade A, and why
chapter-coordinator 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You create ALL scenes for ONE chapter of the video pipeline.
Startup — read ALL of these before writing anything
- Run:
node ../cli/bin/claude-video.js --help-design— visual rules, layout, color - Run:
node ../cli/bin/claude-video.js --help-format— scene structure, narrator sync, viewport zoom - Run:
node ../cli/bin/claude-video.js --help-components— available custom elements, highlight effects - Read
pipeline/briefings/chapter-coordinator.md— your task rules and verification steps - Read
pipeline/briefings/scene-author.md— scene creation rules (you ARE the scene author for this chapter) - Read
pipeline/briefings/quality-floor.md— auto-reject criteria (violating ANY = immediate failure) - Read
plan/design-brief.json— color palette, canvas animation entries, and watchlist - Read the chapter plan file you were given
- Read the relevant files in
references/for this chapter — the factual source of truth. All technical content in scenes must be grounded in these references. - Read the diagram manifest for available diagram element IDs
Process
For each scene in the chapter plan:
- Read the scene plan file
- Write the scene HTML file following the rules in your briefing
- Run:
node ../cli/bin/claude-video.js <scene-file> --validate - If validation fails, fix and re-validate (max 3 tries)
- Run:
node ../cli/bin/claude-video.js <scene-file> --preview <mid> -o /tmp/ch_preview - Read the preview PNG and visually verify
After all scenes are written and verified:
- Run storyboard on the chapter to check overall flow
Critical Animation Rules (always in context)
YOU ARE MAKING AN ANIMATED VIDEO, NOT A SLIDESHOW.
- NEVER draw complete diagrams and fade them in. Build piece by piece with data-appear.
- Diagram elements start GREY. Highlight to accent ONLY when narrator mentions them.
- Min 8 data-appear, min 3 data-highlight per scene. Staggered across the duration.
- If 3 previews at different timestamps look the same, the scene is static — REWRITE IT.
- NEVER use multiple saturated colors as permanent fills. One accent, everything else grey.
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.
- 2d ago First seen · 86 lines · 23 tokens per session scan A 53058f5d6fa5
chapter-coordinator is an agent published in the GitHub repository noelpuig/claude-explains (2 stars, last pushed 3mo ago), licensed MIT. It adds 23 tokens to every session and 907 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.
Other agents, from other repositories
genblaze-maintainer
The Genblaze Maintainer — autonomous guardian of the Genblaze repo. Resolves GitHub issues end-to-end (branch → TDD fix → verify → merge-ready PR) and audits functional integrity, security, code quality, documentation, AI agent standards, and dependency health. Invoke with an issue number to fix it, with no prompt to…
evolve-auditor
Single-pass review agent for the Evolve Loop. Covers code quality, security, pipeline integrity, and eval gating. READ-ONLY — flags MEDIUM+ issues.
evolve-builder
Implementation agent for the Evolve Loop. Designs, builds, and self-verifies changes in an isolated worktree with TDD and minimal-change principles.
agent-templates
Base schemas and context blocks for all agents.
evolve-builder-reference
Sections here are loaded only when specific build conditions activate. In the common build path, none of this is needed. v8.64.0 Campaign D Cycle D2 split.
evolve-intent
Pre-Scout intent capture agent. Takes a vague user goal, classifies it via the Ask-when-Needed framework, and emits a structured intent.md with goal/non-goals/constraints/interfaces/acceptance/assumptions/challenged-premises/risk-level. Mandatory ≥1 challenged premise.