Repos / Eval Kai / Security / AI Fix Prompts
⚡

AI Fix Prompts for Eval Kai

Copy any prompt below into Claude, ChatGPT, or your AI coding assistant to automatically fix the issue. Each prompt includes full context, code location, and step-by-step fix instructions.

4
Total Prompts
0
Critical (P0)
1
High (P1)
3
Medium (P2)
0
Low (P3)
Download All (Markdown) Download All (JSON) Feed these prompts to any AI coder: Claude Code, Cursor, Copilot, GPT, Ollama
HIGH ⚙ moderate #1

Fix quality gate failures (1 conditions)

quality-gate quality
Expected outcome: All quality gate conditions pass
Files to modify: Will be determined by the AI
Prompt (copy this into your AI assistant) 
Repository 'duonghaijoe__eval-kai' is failing the quality gate.

Failed conditions:
- overall_score: actual 42.7 >= 50 (FAILED)

Fix each failing condition to make the repo pass the quality gate.
Methodology: Repobility · https://repobility.com/research/state-of-ai-code-2026/
MEDIUM ⚡ quick-fix #2

Add a LICENSE file

license legal
Expected outcome: LICENSE file added
Files to modify: Will be determined by the AI
Prompt (copy this into your AI assistant) 
Repository 'duonghaijoe__eval-kai' has no detectable license. Add a LICENSE file. For open-source projects, MIT is recommended unless dependencies require a specific license. Create the full file.
MEDIUM ⚙ moderate #3

Fix 5 scorecard failures (50%)

scorecard compliance
Expected outcome: Scorecard score improved from 50% to 80%+
Files to modify: Will be determined by the AI
Prompt (copy this into your AI assistant) 
Repository 'duonghaijoe__eval-kai' fails 5 scorecard checks (score: 50%):

- Has LICENSE: Repository has a LICENSE file
- Has CI/CD: Repository has CI/CD configuration
- Has Tests: Repository has test files
- Quality Gate Passed: Passes the default quality gate
- Grade C or Above: Overall quality grade is C or better

Fix each failing check.
MEDIUM ⚒ significant #4

Simplify 5 high-complexity files

complexity refactoring quality
Expected outcome: All listed files reduced to medium or low complexity
Files to modify: scripts/kai_actor.py, web/frontend/src/components/EnvironmentSettings.jsx, web/frontend/src/components/SessionDetail.jsx, web/frontend/src/components/SessionList.jsx, web/frontend/src/components/Reports.jsx
Prompt (copy this into your AI assistant) 
These files in 'duonghaijoe__eval-kai' have high cyclomatic complexity:

- **scripts/kai_actor.py**: complexity=118, max nesting=6, longest function=87 lines
- **web/frontend/src/components/EnvironmentSettings.jsx**: complexity=99, max nesting=5, longest function=141 lines
- **web/frontend/src/components/SessionDetail.jsx**: complexity=82, max nesting=8, longest function=190 lines
- **web/frontend/src/components/SessionList.jsx**: complexity=80, max nesting=6, longest function=201 lines
- **web/frontend/src/components/Reports.jsx**: complexity=75, max nesting=6, longest function=324 lines

For each file:
1. Break large functions into smaller, focused functions
2. Reduce nesting depth (extract early returns, use guard clauses)
3. Simplify conditional logic
4. Extract complex expressions into named variables