feat(skills): introduce owasp-ml

[JasonTheDeveloper]

Pull Request

Description

In alignment with phase 2 discussed in #480 (comment), this PR introduces the OWASP ML Top 10 skill to hve-core and the security reviewer agent.

Related Issue(s)

Closes #1205

Type of Change

Select all that apply:

Code & Documentation:

• Bug fix (non-breaking change fixing an issue)
• New feature (non-breaking change adding functionality)
• Breaking change (fix or feature causing existing functionality to change)
• Documentation update

Infrastructure & Configuration:

• GitHub Actions workflow
• Linting configuration (markdown, PowerShell, etc.)
• Security configuration
• DevContainer configuration
• Dependency update

AI Artifacts:

• Reviewed contribution with prompt-builder agent and addressed all feedback
• Copilot instructions (.github/instructions/*.instructions.md)
• Copilot prompt (.github/prompts/*.prompt.md)
• Copilot agent (.github/agents/*.agent.md)
• Copilot skill (.github/skills/*/SKILL.md)

Note for AI Artifact Contributors:

• Agents: Research, indexing/referencing other project (using standard VS Code GitHub Copilot/MCP tools), planning, and general implementation agents likely already exist. Review .github/agents/ before creating new ones.
• Skills: Must include both bash and PowerShell scripts. See Skills.
• Model Versions: Only contributions targeting the latest Anthropic and OpenAI models will be accepted. Older model versions (e.g., GPT-3.5, Claude 3) will be rejected.
• See Agents Not Accepted and Model Version Requirements.

Other:

• Script/automation (.ps1, .sh, .py)
• Other (please describe):

Testing

To be able to test the owasp-ml skill using the security reviewer agent you will need a repository containing a mcp code (not configurations).

1. Either select the Security Reviewer agent or invoke the agent via the /security-revew instruction
2. Use the following prompt analyse the code and produce a vulnerability report
   • If you are testing to see if the codebase-profiler.agent.md picks up that the repository contains mcp and thus uses the owasp-ml skill then that's all you need.
   • If you only want to test the owasp-ml is used, in your prompt add targetSkill=owasp-ml

You should see in the output report the owasp-ml skill being referenced and used.

Checklist

Required Checks

• Documentation is updated (if applicable)
• Files follow existing naming conventions
• Changes are backwards compatible (if applicable)
• Tests added for new functionality (if applicable)

AI Artifact Contributions

• Used /prompt-analyze to review contribution
• Addressed all feedback from prompt-builder review
• Verified contribution follows common standards and type-specific requirements

Required Automated Checks

The following validation commands must pass before merging:

• Markdown linting: npm run lint:md
• Spell checking: npm run spell-check
• Frontmatter validation: npm run lint:frontmatter
• Skill structure validation: npm run validate:skills
• Link validation: npm run lint:md-links
• PowerShell analysis: npm run lint:ps
• Plugin freshness: npm run plugin:generate

Security Considerations

• This PR does not contain any sensitive or NDA information
• Any new dependencies have been reviewed for security issues
• Security-related scripts follow the principle of least privilege

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.62%. Comparing base (07bd2ab) to head (988e0b9).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1227      +/-   ##
==========================================
- Coverage   87.63%   87.62%   -0.02%     
==========================================
  Files          61       61              
  Lines        9328     9328              
==========================================
- Hits         8175     8174       -1     
- Misses       1153     1154       +1     

Flag	Coverage Δ
pester	85.18% <ø> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.
see 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[WilliamBerryiii]

@JasonTheDeveloper - how do you anticipate this particular skill is used? I get the other ones for review and evaluation of existing code and for reviews of PRs, but this feels more like a planning setup. My concern with planning is that skills tend to not be great for overview/breadth needs of planning vs deep task specific work where skills tend to excel.

[JasonTheDeveloper]

@WilliamBerryiii I was thinking this skill in particular would help data scientists as they're writing notebooks, for example to train and utilise models. I do see where you're coming from. Although I believe the utilisation of this skill in particular would be a lot lower than others, I felt it might still be relevant. Happy to abandon the PR if you feel it's not necessary.

[WilliamBerryiii]

To the other maintainers ... I'm still thinking through how we provide this capability into the ecosystem. I am of the opinion that it probably makes the most sense as a "planner" system, like SSSC, RAI, and Security (and the forthcoming "accessibility" one). @C-Neisinger ... would also appreciate some input from you on this one.

[chaosdinosaur]

Thank you for introducing the OWASP ML Top 10 skill, @JasonTheDeveloper. The structure and integration are solid — skill validation passes, all agent/prompt/collection references are consistently updated, and the 10 vulnerability reference documents follow a clean, consistent format.

Four inline comments are attached. The highest-priority item is the missing Third-Party Attribution section in SKILL.md — all sibling OWASP skills include this, and it is required by the CC-BY-SA-4.0 license.

RI-05 — PR checklist items (Info): Several AI artifact checklist items are unchecked (Documentation updated, /prompt-analyze review, prompt-builder feedback, common standards verification). Consider running /prompt-analyze on the new skill assets before merge.

Overall this is a well-structured contribution. Looking forward to the updates!

[chaosdinosaur]

RI-01 — Missing Third-Party Attribution (High)

All sibling OWASP skills (owasp-top-10, owasp-llm, owasp-agentic) include a ## Third-Party Attribution section acknowledging OWASP Foundation copyright and CC-BY-SA-4.0 license terms. This section is missing from owasp-ml.

Since the content is derived from the OWASP ML Security Top 10, proper attribution is required by the CC-BY-SA-4.0 license.

Suggested addition (before the closing ---):

## Third-Party Attribution

Copyright © OWASP Foundation.
OWASP® Machine Learning Security Top 10 (2023) content is derived from works by the
OWASP Foundation, licensed under CC BY-SA 4.0
(<https://creativecommons.org/licenses/by-sa/4.0/>).
Source: <https://owasp.org/www-project-machine-learning-security-top-10/>
Modifications: Vulnerability descriptions restructured into agent-consumable reference
documents with added detection and remediation guidance.
OWASP® is a registered trademark of the OWASP Foundation. Use does not imply endorsement.

[chaosdinosaur]

RI-02 — Hyphen instead of em dash (Low)

Lines 27–29 use em dash (—) between the bold skill name and description. This new entry uses a plain hyphen (-), breaking the formatting pattern.

Suggested change

	- **OWASP ML Top 10** - OWASP Machine Learning Top 10 (2023) vulnerability knowledge base for identifying, assessing, and remediating security risks in machine learning systems
	- **OWASP ML Top 10** — OWASP Machine Learning Top 10 (2023) vulnerability knowledge base for identifying, assessing, and remediating security risks in machine learning systems

[chaosdinosaur]

RI-03 — "index" vs "master index" (Low)

All three sibling OWASP skills say master index of all vulnerability identifiers here. Minor wording inconsistency.

Suggested change

	* `00-vulnerability-index.md` — index of all vulnerability identifiers, categories, and cross-references.
	* `00-vulnerability-index.md` — master index of all vulnerability identifiers, categories, and cross-references.

[chaosdinosaur]

RI-04 — Technology signals may be too narrow (Medium)

Other skill signal blocks have 3–5 entries for broader detection coverage. ML codebases commonly include framework imports (torch, tensorflow, sklearn, keras, transformers) and additional model formats (.safetensors). Adding at least one framework-import signal would improve the codebase profiler's detection accuracy.

Consider expanding to:

owasp-ml:
  - "ML training/inference code"
  - "Model files (.pt, .h5, .onnx, .pkl, .safetensors)"
  - "ML framework imports (torch, tensorflow, sklearn, keras, transformers)"