Skill 109
v1.0.0Expertise in deploying, monitoring, detecting drift, automating retraining, and ensuring fairness and compliance for production ML models.
⭐ 0· 205·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (MLOps & Model Governance) matches the SKILL.md content: deployment patterns, versioning, feature stores, drift detection, retraining pipelines, monitoring, and governance. There are no unrelated required binaries, env vars, or config paths that would contradict the stated purpose.
Instruction Scope
The SKILL.md contains high-level, domain-appropriate instructions and example snippets (data quality checks, KS test for drift, automated retraining pipeline, canary rollout). These are prescriptive but remain within MLOps scope. Note: the document uses placeholder helper calls (e.g., load_recent_features, trigger_retraining, alert) without implementation details; if someone implements or runs these, they should verify authentication, rate limits, and safeguards to avoid unintended automated retraining or data access.
Install Mechanism
No install spec is provided and no code files beyond SKILL.md and a harmless package.json are present, so nothing will be downloaded or written to disk by an installer. This is the lowest-risk install profile (instruction-only).
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. The SKILL.md does not instruct the agent to access secrets or unrelated external services. This is proportionate to an advisory MLOps skill.
Persistence & Privilege
Skill flags are default (always: false, user-invocable: true). The skill does not request permanent presence or modify other skills' configuration. Autonomous invocation is permitted by platform default but not additionally privileged here.
Assessment
This skill is a high-quality, instruction-only MLOps guide and appears coherent with its purpose. Before using it in production: (1) treat the code snippets as illustrative—do not run them verbatim without review and testing; (2) implement automation triggers (retraining, canary promotion, alerts) with strict access controls, rate limits, and approval gates to avoid accidental cost or data exposure; (3) ensure any real implementations of helper calls (load_recent_features, alert, trigger_retraining) authenticate to data stores with least privilege and log actions; (4) review any concrete pipeline code you build from this guide for data handling, PII leakage, and compliance requirements; and (5) if you plan to allow the agent to execute these steps autonomously, restrict credentials and review audit logs to mitigate risk.Like a lobster shell, security has layers — review code before you run it.
latestvk971yck43bmcx6g212yg52rf3n82hzr3
License
MIT-0
Free to use, modify, and redistribute. No attribution required.