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 skills add kalyvask/winning-writing --skill bluf-rewritergit clone --depth 1 https://github.com/kalyvask/winning-writingWrote 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/skills/kalyvask/winning-writing/bluf-rewriter)<a href="https://agentmods.dev/skills/kalyvask/winning-writing/bluf-rewriter"><img src="https://agentmods.dev/badge/skills/kalyvask/winning-writing/bluf-rewriter/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/skills/kalyvask/winning-writing/bluf-rewriter"><img src="https://agentmods.dev/badge/skills/kalyvask/winning-writing/bluf-rewriter.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 | $0.00098 | $0.00990 |
| Opus 5 | $0.00049 | $0.00495 |
| Sonnet 5 | $0.00020 | $0.00198 |
| Haiku 4.5 | $0.00010 | $0.00099 |
Grade A, and why
bluf-rewriter 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 9d 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
BLUF rewriter
Source: points/frameworks.md (BLUF section), points/core-rules.md (rule 3).
The rule
Bottom line up front. The conclusion goes in the first sentence. Context, evidence, and qualification come after.
This is a military principle. It applies to every form of business writing: emails, memos, status updates, board reports, performance reviews, investor updates, even op-ed openings.
What "organ music" looks like
"Over the past quarter, we evaluated several scenarios with input from sales, finance, and ops, and after extensive analysis with the team, we have come to believe that…"
40 words of throat-clearing before the actual point. The reader has already started skimming.
What BLUF looks like
"We need to lower Q3 sales forecast by 50%. Three reasons:"
1. Enterprise pipeline collapsed in August (down 60% MoM) 2. Top customer renewed for 1 year instead of 3 3. Competitor launched at 40% lower price point
The conclusion is the first sentence. Evidence follows. Reader knows in 5 seconds whether to keep reading.
How to do it
Step 1: find the actual bottom line
Read the draft. What is the one sentence the reader needs to walk away with? Often it's buried in paragraph 3 or in the conclusion.
Step 2: move it to sentence 1
Cut, paste, done. If it doesn't work as the first sentence, the sentence isn't sharp enough — rewrite it until it does.
Step 3: structure what follows
Choose one of three structures:
A. The "three reasons" memo
- Sentence 1: bottom line
- Sentence 2: "X reasons:" or "Here's why:"
- Bullets
B. The exec update
- Paragraph 1: bottom line + the one number that matters
- Paragraph 2: what's working
- Paragraph 3: what's not
- Paragraph 4: the ask or next step
C. The decision memo
- Sentence 1: the recommendation
- Section: context (1 paragraph max)
- Section: options considered
- Section: why this option
- Section: risks and mitigations
- Section: next steps
Step 4: cut the throat-clearing
Delete every sentence that:
- Summarizes what you're about to say ("In this memo I'll argue…")
- Recaps what the reader already knows
- Builds up to the point ("organ music")
- Hedges the conclusion ("we tentatively believe that perhaps…")
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.
- 9d ago First seen · 102 lines · 98 tokens per session scan A 4e800ca59421
bluf-rewriter is a skill published in the GitHub repository kalyvask/winning-writing (13 stars, last pushed 4d ago), licensed MIT. It adds 98 tokens to every session and 990 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
slide-deck
When you want to draft, update, convert, or export a slide deck for a React/Next.js slide system (${SLIDEDECKREPO:-$HOME/code/your-slide-deck-site}/src/app/slides/). Writes TypeScript Slide[] arrays using your primitives (Eyebrow, Heading, Accent, Body, BulletList, Divider, TwoCol, GradientText), 12 cycling brand…
companion-skill
An AI companion skill for developers and office workers during long-running or repetitive tasks. It provides conversation, feedback, generated lifestyle images, and messages sent through Feishu, a workplace collaboration app.
transcript-cleanup
Cleans and formats a raw auto-generated podcast transcript for publication on a website or show page — removing filler words and transcription errors, adding speaker labels, formatting for readability, and preserving the authentic voice of host and guests.
reels-scripting
Scripts 15–60 second short videos (Reels, TikTok, Shorts) in a timestamped two-column VISUAL | AUDIO/VO format — verbal + visual hook in the first two seconds, a retention beat every 3–5 seconds, pattern interrupts at drop points, on-screen captions, a loop-friendly ending, and a CTA — two script variants per idea.…
youtube-thumbnail
Designs and tests YouTube thumbnails — 3–5 concept briefs covering composition, face or object, a 3-word-max overlay, and colour contrast against YouTube's UI, each run through a curiosity check, a mobile legibility test at 168×94 px, and a /10 scoring rubric, ending in a verdict and an A/B rotation plan. Generates…
agency-multi-brand-pack
Per-client asset templates scoped by workspace.