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 makifbaysal/tasktrooper --skill ml-problem-framinggit clone --depth 1 https://github.com/makifbaysal/tasktrooperWrote 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/makifbaysal/tasktrooper/ml-problem-framing)<a href="https://agentmods.dev/skills/makifbaysal/tasktrooper/ml-problem-framing"><img src="https://agentmods.dev/badge/skills/makifbaysal/tasktrooper/ml-problem-framing/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/makifbaysal/tasktrooper/ml-problem-framing"><img src="https://agentmods.dev/badge/skills/makifbaysal/tasktrooper/ml-problem-framing.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 | $0.00046 | $0.01326 |
| Opus 5.5 | $0.00018 | $0.00530 |
| Sonnet 5.5 | $0.00009 | $0.00265 |
| Haiku 4.5 | $0.00005 | $0.00133 |
Grade A, and why
ml-problem-framing 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 4d 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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ML Problem Framing
Overview
Most failed models were well trained on the wrong question: a target that leaks the outcome, a metric nobody acts on, a population that differs from the one scored in production, or no baseline to show the model is worth its upkeep. Framing is a short written spec, done before the first line of model code.
Core principle: a model exists to change a decision. If you cannot name the decision, the cost of each kind of error and the baseline the model must beat, you are not ready to train.
1. The framing card
Fill this in from the task, the code and the data. Put it in your closing message, and in the repository's model card or README section when the repo keeps one.
| Field | Question | Example |
|---|---|---|
| Decision | What action changes based on the output? | Send a retention offer to the top 5% risk each Monday |
| Target | Exact definition, as code or SQL | churned = no paid order in the 60 days after snapshot_date |
| Unit | One row is one what? | customer × weekly snapshot |
| Prediction time | When is the prediction made, what is known then? | Monday 00:00 UTC, data up to Sunday 23:59 |
| Horizon / label window | How far ahead, and when is the label final? | 60 days; labels final 60 days after snapshot |
| Population | Who is scored, who is excluded? | Active paying customers; exclude staff and test accounts |
| Metric | Which number decides, tied to error costs? | Precision@5% (offer budget is fixed) |
| Baseline | What it must beat | "days since last order" rule; current heuristic |
| Constraints | Latency, interpretability, fairness, retrain cadence | Weekly batch; reasons per customer for the CRM team |
Anything in this table the task does not settle and the data cannot answer — what counts as churn, which population — is a product question: numbered questions via add_task_comment, then stop on it. Choosing between two reasonable technical options (a 60- vs 90-day window when the task said "about two months") is a decision you make and record.
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.
- 4d ago First seen · 91 lines · 46 tokens per session scan A b5228f2c0ca9
ml-problem-framing is a skill published in the GitHub repository makifbaysal/tasktrooper (113 stars, last pushed yesterday), licensed Apache-2.0. It adds 46 tokens to every session and 1,326 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 2026-10-09.
Other skills, from other repositories
flow-nexus-neural
Train and deploy neural networks in distributed E2B sandboxes with Flow Nexus.
agent-v3-memory-specialist
Agent skill for v3-memory-specialist - invoke with $agent-v3-memory-specialist.
agent-data-ml-model
Agent skill for data-ml-model - invoke with $agent-data-ml-model.
agent-neural-network
Agent skill for neural-network - invoke with $agent-neural-network.
agent-safla-neural
Agent skill for safla-neural - invoke with $agent-safla-neural.
agent-sona-learning-optimizer
Agent skill for sona-learning-optimizer - invoke with $agent-sona-learning-optimizer.