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 randommonicle/claude-skills --skill ai-surface-disciplinegit clone --depth 1 https://github.com/randommonicle/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/randommonicle/claude-skills/ai-surface-discipline)<a href="https://agentmods.dev/skills/randommonicle/claude-skills/ai-surface-discipline"><img src="https://agentmods.dev/badge/skills/randommonicle/claude-skills/ai-surface-discipline/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/randommonicle/claude-skills/ai-surface-discipline"><img src="https://agentmods.dev/badge/skills/randommonicle/claude-skills/ai-surface-discipline.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 | $0.00095 | $0.02212 |
| Opus 5 | $0.00048 | $0.01106 |
| Sonnet 5 | $0.00019 | $0.00442 |
| Haiku 4.5 | $0.00010 | $0.00221 |
Grade A, and why
ai-surface-discipline 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 — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI-surface discipline
Every surface that sends data to an LLM inherits the same discipline before it ships. Three pillars (input minimisation, output discipline, human-in-the-loop gate) plus a credential and retention posture. This is a regulatory requirement, not a style preference: the surfaces touch leaseholder personal data and service-charge figures, which sit under UK GDPR, RICS client-confidentiality duties, and LTA 1985 accuracy obligations.
When this applies
This skill fires when building or editing any path that sends data to a model:
- Edge Functions that call an LLM provider
- Draft-commentary, summary, or narrative generators (variance, LPE, FME, year-end)
- Auto-resolvers and suggestion engines, including the reconciliation cash-allocation surface
- Classification, extraction, or matching calls
- Any new AI-assist feature, or any change to an existing one that alters what is sent or how output is used
It does not fire on read-only analytics or queries that never leave the database, or on features that process data entirely in-house with no model call.
Pillar 1: input minimisation
Defence starts at the boundary. The model receives the least data that does the job.
- Hardcoded never-send allowlist. Only named, reviewed fields cross the boundary. The allowlist lives in code, not config, so a data change cannot widen it silently. Default deny: a field not on the allowlist does not leave.
- Redact then restore. Where free text must be sent, redact personal identifiers before the call and restore the tokens in the response, so the user still gets readable output without the PII ever leaving the boundary.
- No PII passthrough. Leaseholder names, contact details, and account identifiers are not sent unless the allowlist names them and the lawful basis is recorded.
- Minimise by Article 5(1)(c). Data minimisation is a UK GDPR principle, not a nicety. If a field is not needed for the task, it is not sent.
- Guard against prompt injection. Free-text fields (dispute notes, maintenance descriptions, anything a leaseholder or third party can write) are data, not instructions. Do not concatenate user-supplied text into an instruction-bearing position, and sanitise it before it reaches the prompt, so a crafted input cannot redirect the model or pull allowlisted data into the response.
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 · 127 lines · 0 tokens per session scan A 3932c1d2010d
ai-surface-discipline is a skill published in the GitHub repository randommonicle/claude-skills (23 stars, last pushed 5d ago), licensed Apache-2.0. It adds 95 tokens to every session and 2,212 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 2026-08-31.
Other skills, from other repositories
llm-app-patterns
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
prompt-optimization
Improve a prompt on the evaluations workbench through a measured loop. Score the baseline first, then duplicate the target column, form a hypothesis from failing rows, edit the copy's prompt draft, run, compare pass rate and cost, and repeat until the numbers hold. Use when the user asks to optimize or improve a…
enhance-prompt
Transforms vague UI ideas into polished, Stitch-optimized prompts. Enhances specificity, adds UI/UX keywords, injects design system context, and structures output for better generation results.
prompt-engineer
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot…
seedance-vocab-en
This skill should be used when an English Seedance 2.0 prompt needs clearer production wording, less generic prose, or precise vocabulary for camera, lighting, motion, VFX, audio, and constraints. Route blocked prompts through seedance-filter for context and boundary review.
ideogram4
Prompting patterns for Ideogram 4 text-to-image — best-in-class in-image text rendering and exact color/layout control via structured JSON captions. Use when generating images that need legible on-image text (title cards, thumbnails, logos, signage, CTAs), precise brand colors, or controlled spatial layout. Triggers…