CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

model-deployment

Deploy ML models with FastAPI, Docker, Kubernetes. Use for serving predictions, containerization, monitoring, drift detection, or encountering latency issues, health check failures, version conflicts.

Quality

79%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./plugins/model-deployment/skills/model-deployment/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

67%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-structured, code-heavy overview with a clear quick start, troubleshooting patterns, and correctly signaled one-level-deep references. Its main weaknesses are redundancy (the Dockerfile, health endpoint, and predict endpoint each appear twice) and a few non-executable fragments (stub monitor class, undefined objects, Python inside a bash block, a missing validation script).

Suggestions

Consolidate the duplicated Dockerfile: keep one canonical copy (the 'Docker Deployment' section) and have the Quick Start heredoc reference it instead of repeating it; do the same for the /health endpoint repeated in Known Issues #1.

Fix the Quick Start code fence: steps 1-2 contain bare Python ('import joblib', 'joblib.dump(...)') inside a ```bash block — move them to a python block or use 'python -c' so the block is actually runnable.

Replace the non-executable fragments: either implement ModelMonitor.detect_drift and define 'model_store'/'alert_manager', or defer entirely to references/model-monitoring-drift.md; and remove or bundle the missing 'scripts/validate_model.py' referenced in the CI/CD snippet.

DimensionReasoningScore

Conciseness

The body is mostly lean code rather than prose, but it repeats the same Dockerfile (full 'Docker Deployment' section plus a heredoc copy in 'Quick Start'), duplicates the /health endpoint (server section plus Known Issues #1), and repeats the predict endpoint, so it could be meaningfully tightened.

3 / 5

Actionability

Most sections are copy-paste ready (FastAPI app, Dockerfile, kubectl rollout/rollback commands, resource limits YAML), but there are gaps: bare Python statements inside a ```bash block in Quick Start steps 1-2, a ModelMonitor stub whose detect_drift is 'pass', undefined names ('model_store', 'alert_manager'), and a referenced scripts/validate_model.py that does not exist in the bundle.

4 / 5

Workflow Clarity

The 6-step Quick Start gives a clear executable sequence with checkpoints ('kubectl rollout status', local docker run test) plus a deployment checklist and rollback procedures, but the quick-start path itself lacks an explicit model validation gate before production rollout — a minor validation gap.

4 / 5

Progressive Disclosure

A dedicated 'When to Load References' section clearly signals all four one-level-deep reference files (verified to exist with substantive content and no further nesting), but roughly 130 lines of Known Issues detail and the duplicated deployment code sit inline in an already ~300-line overview, which is content that could live in the references.

4 / 5

Total

15

/

20

Passed

Description

83%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description in third person that pairs concrete capabilities with an explicit 'Use for...' trigger clause covering both tasks and failure symptoms. It falls just short of top marks on specificity, trigger synonyms, and distinctiveness because the broader MLOps vocabulary and unique differentiators of the body content are absent.

DimensionReasoningScore

Specificity

The description names concrete actions and stack ('Deploy ML models with FastAPI, Docker, Kubernetes', 'serving predictions, containerization, monitoring, drift detection'), but omits capabilities the body actually covers (rollback, versioning, A/B testing, CI/CD), so coverage has minor gaps rather than being comprehensive.

4 / 5

Completeness

It clearly answers both what ('Deploy ML models with FastAPI, Docker, Kubernetes. Use for serving predictions, containerization, monitoring, drift detection') and when ('or encountering latency issues, health check failures, version conflicts'), matching the anchor-5 pattern of concrete capability plus explicit trigger phrases.

5 / 5

Trigger Term Quality

Natural trigger phrases like 'serving predictions', 'drift detection', 'latency issues, health check failures, version conflicts' would match real user requests, but common synonyms such as 'MLOps', 'productionize', or 'model serving' are missing.

4 / 5

Distinctiveness Conflict Risk

The ML-deployment niche with drift detection and model serving triggers is mostly distinct, but Docker/Kubernetes/containerization are broadly shared with general DevOps deployment skills, creating minor overlap risk rather than the minimal risk of a fully distinct niche.

4 / 5

Total

17

/

20

Passed

Validation

87%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

14

/

16

Passed

Repository

secondsky/claude-skills

Path

plugins/model-deployment/skills/model-deployment/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.