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 uk-agents/uk-legal-plugins --skill exam-forecastgit clone --depth 1 https://github.com/uk-agents/uk-legal-pluginsWrote 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/uk-agents/uk-legal-plugins/exam-forecast)<a href="https://agentmods.dev/skills/uk-agents/uk-legal-plugins/exam-forecast"><img src="https://agentmods.dev/badge/skills/uk-agents/uk-legal-plugins/exam-forecast/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/uk-agents/uk-legal-plugins/exam-forecast"><img src="https://agentmods.dev/badge/skills/uk-agents/uk-legal-plugins/exam-forecast.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 | $0.00073 | $0.02301 |
| Opus 5 | $0.00036 | $0.01151 |
| Sonnet 5 | $0.00015 | $0.00460 |
| Haiku 4.5 | $0.00007 | $0.00230 |
Grade A, and why
exam-forecast 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 8d 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 — 179 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/law-student-uk:exam-forecast
- Load
~/.claude/plugins/config/uk-legal-plugins/law-student-uk/CLAUDE.md→ module, lecturer, exam format, syllabus. - Apply the workflow below.
- Intake past exams (PDF, paste, or paths). Confirm sample size.
- Analyse each past exam: format, subject coverage, question style, fact-pattern density, recurring traps.
- Cross-exam pattern analysis — what's stable, what varies.
- Combine with current syllabus to produce forecast: subject weights, format, hobby horses, study emphasis.
- Write
~/.claude/plugins/config/uk-legal-plugins/law-student-uk/exam-forecasts/[module]/forecast-[YYYY-MM-DD].md. Framed as weighting heuristic, not prediction.
Purpose
Every lecturer's exam has fingerprints. The same hypo structures recur. The same traps come back. The same subject ratios repeat. Students who have prior exams study smarter; students who don't, study harder. This skill analyses the prior exams you have and surfaces the patterns.
Not magic. A forecast, not a prediction. The skill cannot tell you what's on the exam — it can tell you what's been on past exams and what's likely to recur based on syllabus coverage.
UK exam context
UK law school exams typically fall into three main types:
- Problem questions (hypos): "Advise A" / "Discuss the liability of X" — apply law to facts using IRAC/CILAC structure.
- Essay questions: "Critically evaluate..." / "To what extent..." — analytical or normative discussion of doctrine or policy.
- Mixed format: a combination of problem questions and essays, often with element of choice.
- SQE1-style: multiple-choice, single-best-answer (if module is SQE1 prep).
- Open book vs. closed book: note the format — closed-book exams reward rule-recall; open-book exams reward issue-spotting and application.
The forecast should identify which format type the lecturer has historically used.
Confidence discipline
- Pattern analysis (what subjects appeared, how many questions per topic, how often policy vs. rule-application) — confident where the exams are clearly in front of me.
- Inference about likely emphasis on upcoming exam —
[UNCERTAIN]is the default; these are forecasts, not certainties. Explicitly frame as "based on the [N] past exams you shared, [topic] appeared in [M]. Your upcoming exam may emphasise it, or the lecturer may rotate — use this as a weighting for review time, not a prediction." - If only 1-2 past exams are available, say so explicitly — any pattern inferred from 1 exam is noise.
- If the lecturer is new (no past exams available), skill can't forecast. Say so; fall back to syllabus-based "these are the subjects covered" only.
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.
- 8d ago First seen · 179 lines · 73 tokens per session scan A d8817eb90c60
exam-forecast is a skill published in the GitHub repository uk-agents/uk-legal-plugins (9 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 73 tokens to every session and 2,301 once invoked, about $0.0004 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 2026-09-03.
Other skills, from other repositories
brand-safety-compliance
Scans outbound copy or creator-submitted content for 6 compliance red lines — medical/efficacy, exaggeration/absolute, beauty-permanence, financial/income, fake-scarcity, fake-rapport — and returns a block/warn verdict with fix suggestions. Use when the user asks to check, scan, review, or vet copy before sending, or…
ask
Answer a question from a finished study, citing back into its evidence, and say plainly when the study does not cover it.
compare
Set two term tables against each other — same concept different term, same term different concept, and what each covers that the other lacks.
enablement-course
Assemble an internal AI enablement course for a company's own employees, sized to where they actually are rather than where a vendor deck assumes they are. Produces the session plan, per-session content and exercises, a facilitator guide for a non-expert, and a way to tell whether it worked. Builds on the tech stack…
study-deps
Use when pulling upstream source or cloning a dependency at its installed version to study internals, read code, or create pinned worktrees.
brief
Use when confusion is only “huh,” “wait, what,” “sorry, you lost me,” or “slow down; explain simply and keep it short.”.