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 mostafa-drz/claude-skills --skill which-to-buygit clone --depth 1 https://github.com/mostafa-drz/claude-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/mostafa-drz/claude-skills/which-to-buy)<a href="https://agentmods.dev/skills/mostafa-drz/claude-skills/which-to-buy"><img src="https://agentmods.dev/badge/skills/mostafa-drz/claude-skills/which-to-buy/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/mostafa-drz/claude-skills/which-to-buy"><img src="https://agentmods.dev/badge/skills/mostafa-drz/claude-skills/which-to-buy.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 | $0.00052 | $0.02517 |
| Opus 5.5 | $0.00021 | $0.01007 |
| Sonnet 5.5 | $0.00010 | $0.00503 |
| Haiku 4.5 | $0.00005 | $0.00252 |
Grade A, and why
which-to-buy 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 19d 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 — 199 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Which to buy
Buying something shouldn't take eleven tabs. This skill researches 2–5 products, lines their specs up side by side on one calm page, shows where they actually differ, and puts a recommendation on top that's picked for what matters to you, not for a generic "best". When you want to dig in, every product and question on the page copies a ready-made prompt you paste back into this chat.
It starts automatically when the request matches: "Sony XM6 vs Bose QC Ultra", "which robot vacuum should I get under $500?", "compare these three" plus links. It's not for browsing a whole category (it asks you to narrow to five), and it never buys anything.
One reference, read it every run: research.md (where
specs and prices come from, what to compare, how to pick). Data shape:
evals/example-compare.json. Defaults:
Configuration, at the end of this file.
The flow
Compare progress:
- [ ] 0. Settings: the chat, then saved settings in memory, then Configuration
- [ ] 1. Frame: what, which candidates (2-5), what matters, budget
- [ ] 2. Research: specs and prices, each value with its source (research.md)
- [ ] 3. Choose rows: what differs and what matters, 8-14 rows
- [ ] 4. Pick: for this person's priorities, with the rows that decide it
- [ ] 5. Write compare.json → validate → render → show the page
- [ ] 6. Close: two lines, then stay in the chat for follow-ups
0. Settings
Look in memory for the line starting "which-to-buy settings:" (country, currency, stores to skip, standing priorities such as "quiet matters more than power"). What the user says in this chat wins, then the saved settings, then Configuration.
1. Frame
Work out, from the chat first:
- What is being bought, and the candidates. Use the ones named or linked. If only a category is given ("a robot vacuum"), shortlist 3 from current, well-reviewed options in the budget and say why each made the list. More than five named: ask which to drop, or keep the five closest to the stated priorities and say which were left out.
- What matters: 1–4 priorities in the user's words ("quiet", "battery", "fits carry-on
luggage"). These become
priorities, and each maps to a row. - Budget and country (prices and availability differ by market).
What ships with it
8 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.
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.
- 19d ago First seen · 199 lines · 52 tokens per session scan A 132737c8856b
which-to-buy is a skill published in the GitHub repository mostafa-drz/claude-skills (4 stars, last pushed 20d ago), licensed MIT. It adds 52 tokens to every session and 2,517 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-10-02.
Other skills, from other repositories
skill-builder
Automatically detect source types and build AI skills using Skill Seekers. Use when the user wants to create skills from documentation, repos, PDFs, videos, or other knowledge sources.
ebay-search
Search eBay listings - find items, auctions, deals, and compare prices.
product-image-processor
Download, resize, and remove backgrounds from product images at scale. Use when the user asks to "process product images", batch-download images from the schedule, strip backgrounds, or standardize product photos.
product-configurator-design
Designing build-your-own product configurators (Tesla-style, custom-pricing, plan-builders) with constraint logic, real-time pricing, validation, and save-and-share mechanics. Honest about infinite-options (decision paralysis), canned-bundles-only (no real customization), and guided-configuration (smart defaults plus…
product-photoshoot
Create faithful studio, lifestyle, and on-model product photography through the GooseWorks Product Photos workflow, then approve selected results for reuse in future creative work.
competitor-price-tracker
Monitor competitor pricing pages and send alerts when prices change. Track discount patterns, promotional cycles, and pricing strategy shifts.