agentleFS
Sign inStart free
ForgeyClapForgeyClap/Forge-Core/.claude/skills/forge-mlops/SKILL.md

forge-mlops

Skill2 starsChanged 5 days agoMITView on GitHub

Playbook for production ML: model registry, serving, drift (gebruik bij ML in productie).

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use for: production ML: reproducible training, versioning, train/serve parity, evaluation, drift monitoring.

    line 11

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Forge playbook — Production ML / MLOps

When to use

  • Use for: production ML: reproducible training, versioning, train/serve parity, evaluation, drift monitoring.
  • Domain framing, team roles and honesty rule: references/overview.md.

Do not duplicate ECC skills — defer to: systematic-debugging (pipeline/training failures), /test-coverage (data-validation + feature-code tests), forge-data (the upstream ETL/feature pipeline), forge-prediction (uncertainty labeling + no-auto-bet rules when the model informs a betting/forecast decision). This file is orchestration only.

Hard rules

  • Decision note before training.
  • Metric selection by mistake cost.
  • Reproducible training.
  • Model versioning + working rollback.
  • No train/serve skew.
  • Offline + online evaluation before promotion.
  • Drift + decay monitoring.
  • No automatic real-money / irreversible / outward-facing action — ever.

Full text of every rule: references/hard-rules.md.

Team

Lead: build-boss (or integration-boss when serving wires into external systems). Full roster: references/team-and-tools.md.

Skills / commands / MCP

Skills, commands, MCP and opt-in dependencies: references/team-and-tools.md.

Fan-out & flow

L3 (a real production ML system is inherently multi-component). Parallel: data-validation gate ∥ feature pipeline ∥ offline-eval harness ∥ drift/monitoring scaffolding (independent once the feature contract + metric are fixed). Serial: data validation → feature engineering (versioned, parity-checked) → reproducible training → offline eval vs incumbent → register + version artifact → shadow/canary online eval → monitored rollout → owner-gated promotion. The go-live decision is a human checkpoint, not a pipeline step.

Domain gates

Training reproducible (seed + pinned deps + versioned data/config regenerate the metric); model versioned in a registry with a tested rollback + fallback; no train/serve skew (feature parity proven); data-validation gate at entry; offline eval compares against the incumbent with the right metric + validation scheme; online eval via shadow/canary before any promotion; feature/prediction drift + performance decay monitored with alerts and defined retrain triggers; serving has health checks/timeouts/bounded retries; no auto real-money/irreversible action path exists; no secrets in pipeline code or artifacts.

Ship-readiness

Evidence checklist before handoff: references/ship-readiness.md.

Safety, approval and honesty

  • Approval / safety: promotion to production is owner-gated; no automatic real-money / irreversible / outward-facing action.
  • Approval / safety: no secrets or credentials in pipeline code or model artifacts.
  • Honest failure: no Python/ML env or serving infra: label training, metrics, canary and monitors not-run.
  • Untrusted content: see references/untrusted-content.md.
  • Shared gates for every playbook (honesty, owner approval, secrets, reviews): .claude/docs/forge-reference/playbook-template.md.

References

  • references/overview.md — domain framing paragraph(s) from under the title.
  • references/hard-rules.md — full text of every hard rule.
  • references/team-and-tools.md — full team roster, skills, commands, MCP and opt-in runtimes.
  • references/ship-readiness.md — full ship-readiness evidence checklist.
  • references/untrusted-content.md — Untrusted-content note.
  • references/iteration-and-promotion.md — iteration compact, observation ledger, error-analysis loop, data contract, fail-closed promotion gates.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free