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 reatlat/fullstory-claude-plugin --skill predictive-alertsgit clone --depth 1 https://github.com/reatlat/fullstory-claude-pluginWrote 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/reatlat/fullstory-claude-plugin/predictive-alerts)<a href="https://agentmods.dev/skills/reatlat/fullstory-claude-plugin/predictive-alerts"><img src="https://agentmods.dev/badge/skills/reatlat/fullstory-claude-plugin/predictive-alerts/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/reatlat/fullstory-claude-plugin/predictive-alerts"><img src="https://agentmods.dev/badge/skills/reatlat/fullstory-claude-plugin/predictive-alerts.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 | $0.00047 | $0.01378 |
| Opus 5 | $0.00023 | $0.00689 |
| Sonnet 5 | $0.00009 | $0.00276 |
| Haiku 4.5 | $0.00005 | $0.00138 |
Grade A, and why
predictive-alerts 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 10d 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 — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Predictive Alerts
Set up proactive alerts — warn the user before a metric crosses a critical threshold, not after.
When to Use
- "Alert me if conversion is about to drop below 15%"
- "Warn me if errors are trending toward our SLO limit"
- "Tell me when rage clicks are on track to double this month"
- "Set up a watch on the new checkout flow — alert if anything changes"
- "What should I be worried about tomorrow?"
Mental Model
Reactive monitoring tells you what already happened. Predictive alerting tells you what's about to happen based on the trend.
If error rate is 0.8% today, 1.2% yesterday, and 1.6% the day before — you're on track to hit your 3% SLO limit in 3 days. Alert now, not when you breach it.
Workflow
Step 1: Define the alert
What metric, what threshold, what direction?
Alert: Checkout conversion drops below 15%
Current: 21%
Trend: Declining -0.5pp per week
Predicted threshold breach: ~12 weeks (if trend holds)
→ No immediate alert needed, but set a watch for <18% as early warning
Step 2: Analyze the trend
Build a trend metric and assess the trajectory:
fullstory:build_metric(query="checkout conversion rate", output_type="trend")
fullstory:compute_metric(metric_id, time_range="last_90_days")
Analyze the trend:
- Direction: Up, down, or flat?
- Velocity: How fast is it changing? (per day, per week)
- Acceleration: Is the change speeding up or slowing down?
- Volatility: How much does it bounce around? (noisy trends are harder to predict)
Step 3: Predict threshold breach
Simple linear projection:
Current value: 21%
Weekly change: -0.5pp
Threshold: 15%
Weeks to breach: (21 - 15) / 0.5 = 12 weeks
Caveat heavily: "This is a simple linear projection based on the last 4 weeks. Trends change. This is an early warning, not a forecast."
Step 4: Check for acceleration
Is the decline speeding up?
Last week: -0.3pp
This week: -0.5pp
Week before: -0.2pp
→ The decline is accelerating (from -0.2 to -0.5). If it continues accelerating, the breach could happen in 6-8 weeks, not 12.
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.
- 10d ago First seen · 146 lines · 47 tokens per session scan A fcf6e952c880
predictive-alerts is a skill published in the GitHub repository reatlat/fullstory-claude-plugin (62 stars, last pushed 28d ago), licensed MIT. It adds 47 tokens to every session and 1,378 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
roblox-input
Use when handling Roblox keyboard, mouse, gamepad, touch, motion input, or cross-platform action binding.
roblox-lighting
Use for Roblox lighting, atmosphere, day/night, or post-processing effects.
roblox-performance
Use when profiling Roblox performance or diagnosing FPS, memory, network, mobile, or hot-path problems.
roblox-growth-design
Use for growth diagnosis, discovery, retention, onboarding, experiments, LiveOps, and packaging.
roblox-networking
Use when validating RemoteEvent or RemoteFunction arguments, adding rate limits, designing server-authoritative systems, or preventing exploits.
roblox-ui-design
Use for Roblox UI composition, hierarchy, visual systems, style inheritance, accessibility, and design review.