azure-ai-contentsafety-py

Azure AI Content Safety SDK for Python. Use for detecting harmful content in text and images with multi-severity classification.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Advisory

Suggest reviewing before use

Quality

Content

65%

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 strong, executable API reference with high actionability and tidy organization. Its gaps are the lack of a multi-step workflow with validation, the absence of any referenced supporting files, and a low-value boilerplate tail.

Suggestions

Replace the generic "When to Use"/"Limitations" boilerplate with skill-specific guidance (e.g., when to prefer 8-severity mode, how to choose thresholds) or remove it.

Add a short end-to-end moderation workflow with an explicit validation/feedback checkpoint (analyze -> check severity threshold -> block/halt -> log) so multi-step use is sequenced.

Move the reference tables (Harm Categories, Severity Scale, Client Types) into a REFERENCES.md and signal it from the body to improve progressive disclosure.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean, executable code with compact tables, but the generic "When to Use" and "Limitations" boilerplate ("applicable to execute the workflow or actions described in the overview") earns little relative to its tokens, keeping it below the every-token-earns-its-place anchor.	2 / 3
Actionability	It provides complete, copy-paste-ready code for authentication, text/image analysis, blocklist management, and severity modes with real imports and runnable calls, matching the fully-executable top anchor.	3 / 3
Workflow Clarity	Operations are presented as clearly sectioned independent recipes, but there is no sequencing, error-handling, or validation guidance for combining them into a moderation pipeline, so checkpoints are absent.	2 / 3
Progressive Disclosure	No bundle files exist and all ~210 lines sit inline in a single file; sections are well-organized but the API reference is large enough that it could be split into referenced files, matching the "could be better organized / content that should be separate is inline" anchor.	2 / 3
	Total	9 / 12 Passed

Description

57%

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

The description is specific and distinguishable, correctly stating the niche and capability. Its weakness is trigger coverage and the when-clause, which restates the what rather than offering natural user phrasings.

Suggestions

Add natural trigger terms users actually say, e.g. "content moderation", "profanity/abuse filtering", "text or image moderation".

Expand the when-clause into concrete scenarios, e.g. "Use when moderating user-generated or AI-generated content, screening for hate/self-harm/sexual/violence, or building a safety filter."

Optionally enumerate concrete actions beyond detection, such as classifying severity, applying custom blocklists, and pre-moderating model outputs.

Dimension	Reasoning	Score
Specificity	"detecting harmful content in text and images with multi-severity classification" names the domain and one core action, but does not enumerate multiple distinct concrete actions, so it lands above the vague anchor but below the multi-action top anchor.	2 / 3
Completeness	A "Use for…" clause provides explicit trigger guidance, but it merely restates the capability rather than naming concrete when-scenarios, so the "when" is present yet only implied-level rather than rich.	2 / 3
Trigger Term Quality	It surfaces relevant terms ("harmful content", "text", "images") but omits natural variations a user would say ("content moderation", "profanity filter", "abuse detection"), so coverage is partial.	2 / 3
Distinctiveness Conflict Risk	"Azure AI Content Safety SDK for Python" names a specific, narrowly-scoped product, making it clearly distinguishable and unlikely to trigger for unrelated skills.	3 / 3
	Total	9 / 12 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: boisenoise/skills-collections
Commit: f36337d

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.