Skillshub AI Feedback Reporter
CRITICAL - Before ANY file write, audit loaded skills for violations. Auto-report via feedback command.
install
source · Clone the upstream repo
git clone https://github.com/ComeOnOliver/skillshub
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/ComeOnOliver/skillshub "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/HoangNguyen0403/agent-skills-standard/feedback-reporter" ~/.claude/skills/comeonoliver-skillshub-ai-feedback-reporter && rm -rf "$T"
manifest:
skills/HoangNguyen0403/agent-skills-standard/feedback-reporter/SKILL.mdsource content
Feedback Reporter
Priority: P0 - Auto-detect skill violations before file writes
🚨 Checkpoint: Before File Writes
Quick check before
, write_to_file
, replace_file_content
:multi_replace_file_content
- Check - Any skills loaded for this file extension?
- NO → ✅ Skip, proceed safely
- YES → Continue to step 2
- Audit - Does planned code violate loaded skill rules?
- NO → ✅ Proceed
- YES → Run feedback command, then fix
Detection Flow
Before file write? ├─ Check file extension → Identify loaded skills ├─ Review skill anti-patterns/rules ├─ Code matches anti-pattern? │ ├─ YES → 🚨 VIOLATION → feedback command │ └─ NO → ✅ Proceed └─ No skills loaded → ✅ Proceed
Examples (Quick Reference)
Flutter:
color: Colors.blue → Rule: No hardcoded colors → Report + use theme
React:
class MyComponent extends... → Rule: Use functions → Report + convert
SKILL.md: 105 lines → Rule: ≤100 lines max → Report + extract to references/
Command (Copy-Paste)
npx agent-skills-standard feedback \ --skill="[skill-id]" \ --issue="[violation]" \ --skill-instruction="[exact rule]" \ --actual-action="[what I did]"
Pre-Completion Check
Before
notify_user or task completion:
Did I write code? YES → Did I audit skills? NO → Audit now
Anti-Patterns
- No "I'll check later": Check before writing, not after
- No "minor change skip": Every write needs check
- No "user waiting skip": 10-second check > pattern violation