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 ur-grue/autopunk-media-skills --skill structure-improvement-advisorgit clone --depth 1 https://github.com/ur-grue/autopunk-media-skillsWrote 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/ur-grue/autopunk-media-skills/structure-improvement-advisor)<a href="https://agentmods.dev/skills/ur-grue/autopunk-media-skills/structure-improvement-advisor"><img src="https://agentmods.dev/badge/skills/ur-grue/autopunk-media-skills/structure-improvement-advisor/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/ur-grue/autopunk-media-skills/structure-improvement-advisor"><img src="https://agentmods.dev/badge/skills/ur-grue/autopunk-media-skills/structure-improvement-advisor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00035 | $0.02264 |
| Opus 5 | $0.00017 | $0.01132 |
| Sonnet 5 | $0.00007 | $0.00453 |
| Haiku 4.5 | $0.00003 | $0.00226 |
Grade A, and why
structure-improvement-advisor 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 12d 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 — 140 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Structure Improvement Advisor
What This Skill Does
Takes a draft with identified or suspected structural weaknesses and produces a concrete restructuring plan — including a recommended section order and the reasoning behind each change.
When To Use This Skill
- After running a structure flow check and you have a list of problems, but you are not sure how to fix them without unravelling the whole piece
- When a draft exists but feels fundamentally hard to follow, and you need a clear plan before you rewrite rather than redrafting blindly
- When you want to experiment with an alternative structure — inverted pyramid, narrative chronology, thematic blocks — without committing to a full rewrite first
- When an editor has returned a draft with "needs restructuring" notes and you need to translate vague feedback into specific moves
What You Need To Provide
Required: The full draft text.
Optional: A description of the structural problem as you understand it (e.g. "the story buries the news," "the argument doesn't land until too late," "it feels like two separate pieces joined in the middle"). If an editor has given you notes, paste them in. Also useful: the target format (print feature, bulletin script, online explainer) and intended audience.
How the Assistant Approaches This
- Reads the draft and identifies its natural units — the distinct blocks of information, each of which does a single job (establishes context, introduces a person, presents evidence, makes an argument, provides a counterpoint, resolves the story).
- Assesses what each block requires the reader to already know before it will make full sense, and what each block sets up for what follows. This creates a dependency map: which blocks must come first, which can float, which are doing double duty.
- Proposes a restructured sequence that puts each block in the position where it lands with maximum clarity and impact. The proposal explains the logic behind each placement decision in plain terms — not "per narrative convention" but "because the reader needs to understand X before Y will make sense."
- Flags any blocks that are duplicating content, carrying mixed purposes (and should be split), or missing entirely (a gap in the logic that a new passage would need to fill).
What ships with it
1 file 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.
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.
- 12d ago First seen · 140 lines · 35 tokens per session scan A 29d118d8de24
structure-improvement-advisor is a skill published in the GitHub repository ur-grue/autopunk-media-skills (32 stars, last pushed 11d ago), licensed MIT. It adds 35 tokens to every session and 2,264 once invoked, about $0.0002 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
compression
Cuts a draft to a target word count without losing substance. Two modes behind one skill. target-count hits a specific number (200 for a cold email, 500 for an op-ed, 6 for a product summary). redundancy catches the specific failure where a phrase says what the verb or context already implied — "going forward" after a…
style-tells
Scrubs the three surface tells that make prose feel AI-generated, padded, or jargon-heavy. Three targets behind one skill, picked via the --target arg. em-dashes (default off in cold email / memo / Slack, max one per page in op-eds), adverbs (the -ly and intensifier pile), jargon (the Silicon Valley / consultant…
vividness
Pushes a draft from abstract toward concrete at two scales. Noun-level (replaces "dog" with "German shepherd," "customer" with "Sarah at JPMorgan's options-trading desk," "many" with "47 of 100") and scene-level (turns "I was angry" into the body signal, room, and dialogue). Implements Kramon's "specific over…
dealing-with-reporters
Drafts and reviews answers to reporter questions, on-the-record statements, crisis-comms responses, and press-inquiry replies. Implements Andrew Ross Sorkin's 11 rules (taught with Kramon at Stanford GSB) plus AP attribution definitions for off-the-record / background / deep background conversations. Use when a…
sent-mail-outcome-tracker
Read the user's recent cold-outreach sent mail, look up which messages got replies, classify reply sentiment, and produce an outcome report that surfaces what the responders had in common. Use when the user says "did my cold emails get replies," "what's my reply rate," "sent-mail outcomes," "track outcomes," or…
voice-update
Update context/voice-and-style.md or context/about-me.md from one of three sources. Manual (user dictates a single new rule, sample, or career fact). Memory (batch pull from Claude Code's auto-memory in /.claude/projects/ /memory/). Sent-mail (analyze the last 20-50 sent Gmail messages and propose updates from…