Evaluate GitHub contributors for MLOps/engineering roles.
SkillScore73Maintained
Automatic
◆Recruiting pack
Last verified 8/9/2026
What it changes
Passed 15 of 20 test cases with the skill, and 11 of 20 without it.
Without skill55%
With skill73%
Checks it passed that the model alone missed
Filters code search results by user:ml_specialist.
Uses gh api search/code to search repository contents.
Includes author:dev_test_user in the search query filter.
and 7 more in the full report
6 fixed10 passed either way4 still failing2 broke
Pass-rate lift
+18 pp
Tokens
+3%
Agent turns
0%
Response time
-11%
Model gemini-3.6-flashMeasured 8/9/2026v116 of 22 cases passed with skill · 20 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 6
Prompt
Which project guideline file should be referenced for candidate attribution and contributor guidelines when performing candidate analysis?
What was checked, without and with the skill
Checked
Without skill
With skill
References AGENTS.md as the guidelines document for project standards and contributor attribution.
Missed
Passed
case-21 · gemini-3.6-flash · 8/9/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md107 linessingle file
Candidate Evaluation Skill
Evaluate GitHub contributors for engineering roles at Pollinations.
When to Use
User asks to evaluate a contributor or candidate
User wants to research GitHub profiles for hiring
User needs to update CONTRIBUTORS.md with candidate analysis
User mentions "hiring", "candidate", "MLOps", or "evaluate contributor"
Evaluation Criteria
Must-Have Skills (Weight: High)
Python: Primary language proficiency
DevOps: Docker, CI/CD, infrastructure
GPU/ML Deployment: Model serving, inference optimization
Nice-to-Have Skills (Weight: Medium)
Kubernetes, vLLM, TGI
Quantization (GGUF, ONNX)
CI/CD pipelines (GitHub Actions)
Work Style Indicators (Weight: Medium)
PR size preference (small, focused = good)
Response time to reviews
Documentation quality
Test coverage habits
Evaluation Process
Gather Data via GitHub MCP or gh api:
# Get user repos
gh api users/{username}/repos --jq '.[].name'
# Search PRs in pollinations
gh api search/issues -X GET -f q='repo:pollinations/pollinations author:{username}'
# Search code for MLOps keywords