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 agentmods add skills/mxslr/mlcraft/domain-recommendernpx skills add mxslr/mlcraft --skill domain-recommendergit clone --depth 1 https://github.com/mxslr/mlcraftWrote 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/mxslr/mlcraft/domain-recommender)<a href="https://agentmods.dev/skills/mxslr/mlcraft/domain-recommender"><img src="https://agentmods.dev/badge/skills/mxslr/mlcraft/domain-recommender.svg" alt="Measured on agentmods" 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 | $0.00096 | $0.00523 |
| Opus 5 | $0.00048 | $0.00262 |
| Sonnet 5 | $0.00019 | $0.00105 |
| Haiku 4.5 | $0.00010 | $0.00052 |
Grade A, and why
domain-recommender 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.
What it actually says
Recommendation - Method Selection
For large catalogs use two stages: retrieve candidates, then rank them. Always keep a simple, strong baseline.
Decision table
| Sub-task | Recommended | Notes |
|---|---|---|
| Strong baseline | matrix factorization (ALS, BPR) or item k-NN | cheap and hard to beat. Always include it. |
| Candidate retrieval (large catalog) | two-tower model (user tower and item tower) + approximate nearest neighbor index | scalable first stage. |
| Ranking | gradient-boosted trees on features, or DeepFM or DLRM | optimizes the top of the list. |
| Sequential or next-item | SASRec (causal) or BERT4Rec (bidirectional, cloze objective) | captures order. The bidirectional cloze can leak the target if not masked correctly. |
| Cold start | content-based features inside a two-tower model | for new users or items with no history. |
Cross-cutting practice
- Leakage (critical): split by TIME (train on the past, evaluate on the future) or leave-one-last-interaction-out per user. Never use a random interaction split. Never let the model see the target interaction.
- Handle implicit feedback with BPR or sampled softmax; address popularity bias and the cold-start case.
- Metrics: Recall@K, NDCG@K, MAP, MRR, HitRate@K. Do not use plain accuracy or ROC-AUC as the headline. Always compare against a popularity baseline.
- Explainability: nearest items, "because you interacted with X", and SHAP on the ranker features.
- Recent direction (2024-2025): LLM-based generative recommenders and generative retrieval. SASRec and BERT4Rec remain strong, still-standard baselines; reported gains over them are frequently overstated, so always compare fairly.
- Improve accuracy: use
accuracy-improvement-loop(hard-negative mining, a sequential model, or a stronger two-stage pipeline).
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 · 26 lines · 96 tokens per session scan A 0d5110e347a5
domain-recommender is a skill published in the GitHub repository mxslr/mlcraft (8 stars, last pushed 1mo ago), licensed MIT. It adds 96 tokens to every session and 523 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
seo-strategy
Audits and improves organic search performance — technical health, site architecture, internal linking, structured data, and the content decisions that determine what can rank. Use this to run an SEO audit, diagnose why pages are not ranking or were deindexed, plan a site's URL and navigation structure, add structured…
implement-factory
Factory loop orchestrator for multi-feature or multi-component implementation manifests. Use for high-complexity work with parallel-eligible workstreams and holdout-scenario evaluation.
chief-strategy-officer
Owns where the business plays and how it wins over a multi-year horizon — portfolio choices, corporate development, strategic partnerships, and planning under uncertainty. Use this for a decision about which markets or businesses to be in, whether to build, buy, or partner, how to allocate capital across business…
scenario-planning
Plans under genuine uncertainty — building scenarios, identifying which assumptions are load-bearing, setting early-warning indicators, and stress-testing a plan against futures rather than forecasting one. Use this when a decision depends on something unknowable, when a plan assumes conditions that may not hold…
ai-ml-governance
Governs models and AI systems in production — intended use, evaluation, monitoring, human oversight, documentation, and the decision to deploy or retire. Use this before deploying a model or AI feature, when defining evaluation criteria, when a model's behavior has drifted, when assessing AI risk or regulatory…
ai-research-analyst
Produces executive-level research — market sizing, competitor mapping, trend analysis, and strategic intelligence — grounded in cited sources with the confidence in each claim made explicit. Use this to analyze a market or industry, map competitors, evaluate a market-entry or build-versus-buy decision, produce a…