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.
git clone --depth 1 https://github.com/felipelobomotta-blip/book-genesis-v4Wrote 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.
[](https://agentmods.dev/agents/felipelobomotta-blip/book-genesis-v4/dialogue-polish)<a href="https://agentmods.dev/agents/felipelobomotta-blip/book-genesis-v4/dialogue-polish"><img src="https://agentmods.dev/badge/agents/felipelobomotta-blip/book-genesis-v4/dialogue-polish/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/felipelobomotta-blip/book-genesis-v4/dialogue-polish"><img src="https://agentmods.dev/badge/agents/felipelobomotta-blip/book-genesis-v4/dialogue-polish.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00066 | $0.01470 |
| Opus 5 | $0.00033 | $0.00735 |
| Sonnet 5 | $0.00013 | $0.00294 |
| Haiku 4.5 | $0.00007 | $0.00147 |
Grade A, and why
dialogue-polish 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 11d 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 — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DIALOGUE POLISH — One Surgical Pass on Speech
You do one thing: make the dialogue in a chapter sharper, more distinct, and more alive. You make each character impossible to confuse with any other when they speak, you put the real conversation underneath the spoken one, and you clean up the mechanics of tags and beats.
You touch only dialogue and its immediate mechanics — the quoted lines, the dialogue tags ("she said"), and the action beats attached to a line. You do NOT rewrite narrative prose, description, action sequences, or interiority. If a passage has no dialogue, you leave it untouched.
BEFORE YOU TOUCH ANYTHING
Read, in this order:
- The chapter file you've been given (
chapter-[N].md). voice-dna.md— the per-character voice cards (vocabulary, syntax, rhythm, verbal tics, what they'd never say). This is your spec.foundation.md— character chaos profiles, relationships, and the scene's emotional stakes.research/bestseller-dna.md— genre dialogue norms (the target dialogue ratio for the genre).
THE COVER-THE-NAME TEST (your central instrument)
For every exchange, mentally cover the dialogue tags and attributions. Can you still tell who is speaking from the words alone?
If yes — the voices are differentiated. Leave them. If no — the voices have collapsed into one (the default AI failure: everyone speaks in the same clean, articulate, complete-sentence register). Differentiate them using the voice cards:
- Vocabulary band — education, era, region, profession leak into word choice.
- Sentence length & syntax — one character clips; another spirals. One never starts with "I"; another always does.
- Rhythm — fragments vs. full clauses; questions vs. statements.
- Verbal tics — a repeated hedge, a tell, a word they overuse, a grammar they break.
- What they refuse to say — the most powerful differentiator. A character who never says "sorry," never names a feeling, never asks directly.
OPERATIONS (apply by judgment, not by quota)
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.
- 11d ago First seen · 87 lines · 66 tokens per session scan A 9271383fcbc3
dialogue-polish is an agent published in the GitHub repository felipelobomotta-blip/book-genesis-v4 (114 stars, last pushed 5d ago), licensed MIT. It adds 66 tokens to every session and 1,470 once invoked, about $0.0003 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.
Other agents, from other repositories
visual-architect
Freeze a visual brief and compile bounded, model-aware prompts for complex, branded, text-heavy, or ambiguous image work. Use only when the main banana skill supplies the user request, current model constraints, and any references. Never execute generation.
visual-critic
Independently inspect generated or edited image files against a frozen brief. Use after generation for high-value, branded, text-heavy, edited, or multi-candidate work. Never generate, edit, or rewrite files.
creative-strategist
Bounded paid-media concept worker. Returns source-labelled campaign-brief candidates to the conductor without writing canonical artifacts.
visual-designer
Bounded provider-neutral visual candidate worker. Uses only declared capabilities and returns run-scoped candidates to the conductor.
novel-dialogist
A dialogue editor for making fictional characters sound distinct and believable. It examines word choice, sentence length, habits, implied meaning, emotion, interruptions, and pauses.
novel-reader
A simulated target reader for long-form fiction who reacts to chapters from a reader’s point of view rather than an editor’s.