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 elvisun/newsjack --skill prompt-proximity-architecturegit clone --depth 1 https://github.com/elvisun/newsjackWrote 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/elvisun/newsjack/prompt-proximity-architecture)<a href="https://agentmods.dev/skills/elvisun/newsjack/prompt-proximity-architecture"><img src="https://agentmods.dev/badge/skills/elvisun/newsjack/prompt-proximity-architecture/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/elvisun/newsjack/prompt-proximity-architecture"><img src="https://agentmods.dev/badge/skills/elvisun/newsjack/prompt-proximity-architecture.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 | $0.00067 | $0.01761 |
| Opus 5 | $0.00034 | $0.00881 |
| Sonnet 5 | $0.00013 | $0.00352 |
| Haiku 4.5 | $0.00007 | $0.00176 |
Grade A, and why
prompt-proximity-architecture 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 — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prompt Proximity Architecture
Design the cells before writing the strings.
This skill inherits the ethical floor from skills/ETHICS.md. It enforces anti-hallucination and evidence-bound coverage. Anti-spray and human-send are not applicable.
Inputs
Require:
- a measurement charter;
- approved
icp_hypotheses.jsonandbuyer_jobs.json; - target run/review budget;
- required locales, surfaces, and lanes;
- any campaign partition and prior-panel constraints.
If the charter is provisional, design a provisional architecture and name the gaps. Reject a charter that says only “track AI visibility.”
Keep dimensions independent
Each intent cell fixes:
- buyer job;
- information act;
- journey state;
- material constraints;
- persona or buying role;
- locale and language;
- prompt-proximity band;
- expected answer kind.
Add independent tags for evidence grade, partition, lane eligibility, turn form, and optional funnel. Do not make funnel the schema.
Use the supported acts explain, diagnose, plan, generate, compare, recommend, verify, navigate, buy, implement, and troubleshoot. Include only acts entailed by the job evidence.
Use journey states problem_identification, exploration, requirements_building, supplier_selection, adoption, and post_purchase.
Assign proximity
| Band | Name | Structure | Default aided status |
|---|---|---|---|
B0_direct_brand_product |
Brand | Names target brand/product; asks about facts, fit, use, reputation, support, or implementation | target_aided |
B1_comparison_purchase |
Shortlist | Shortlist, recommendation, alternatives, pricing, requirements, or comparison | unaided; competitor_aided when only competitors are supplied; target_aided for a declared target-vs-competitor comparison |
B2_category |
Category | Names an accepted solution category, not the target | category_aided |
B3_problem_need |
Problem | Describes pain, risk, trigger, or constraint without category/target | unaided |
B4_job_goal |
Goal | Asks for progress/outcome without supplying a solution category | unaided |
B5_broad_discovery_story |
Market | Trend, event, regulation, practice, or narrative connected to the job | unaided |
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 · 170 lines · 67 tokens per session scan A 26ebb9caf3bf
prompt-proximity-architecture is a skill published in the GitHub repository elvisun/newsjack (667 stars, last pushed 10d ago), licensed MIT. It adds 67 tokens to every session and 1,761 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 skills, from other repositories
x402
HTTP 402 payment protocol for AI agent commerce — three-actor model (Client, Resource Server, Facilitator), ERC-3009 transferWithAuthorization, server middleware (@x402/express), client patterns in TypeScript and Python, facilitator integration, agent-to-agent payments, pricing strategies, and replay protection. Works…
splits
Use Splits with Bankr for onchain treasury operations: secure assets, process revenue, manage operating subaccounts, pay expenses, govern contracts, and maintain clean accounting books.
binance-tokenized-securities-info
Query Ondo tokenized US stock data on Binance Web3. Covers: supported stock token list, RWA metadata (company info, attestation reports), market and per-asset trading status (with corporate action codes for earnings, dividends, splits), real-time on-chain data (token price, holders, circulating supply, market cap), US…
payment-assistant
Binance Pay Assistant - Send and Receive crypto payments. Send: QR code payment from Funding Wallet (C2C + PIX). Use when user wants to buy/purchase/pay/transfer/send, confirm/cancel payment, or query order status. Requires QR code data. PIX QR codes (pix, br.gov.bcb.pix) are auto-detected. Receive: Generate QR codes…
binance-wallet-tracker
Wallet tracking and monitoring on Binance Web3. Monitor token/trade activity for private groups or public Smart Money/KOL data. Four domains: (A) Discovery — consensus, pioneer, leaderboard diff, token buyer lookup, time-window summaries, similar wallets; (B) Evaluation — token risk (tokenRiskLevel ≥ 3), group rhythm…
binance-leaderboard
On-chain wallet leaderboard and Gem Hunter analysis on Binance Web3. Query top trader rankings by PnL, win rate, volume, trade count, or token count across BSC/Solana/Base/ETH. Analyze individual wallet addresses with a 6-dimension scoring model (winrate/stability/drawdown/ tags/pnl/followfriendly) plus AI archetype…