1.9 KiB
Forced Eval Skill Hook
Purpose: Force reliable skill activation by creating a commitment mechanism that requires explicit evaluation of all skills before proceeding with implementation.
Success Rate: 84% (vs ~50% baseline)
Source: Based on testing by Scott Spence - https://scottspence.com/posts/how-to-make-claude-code-skills-activate-reliably
MANDATORY SKILL EVALUATION PROTOCOL
Before you proceed with ANY implementation work, you MUST complete this three-step evaluation process:
STEP 1: EVALUATE (MANDATORY - DO THIS NOW)
Go through EACH available skill and explicitly state:
SKILL EVALUATION:
- [skill-name]: YES/NO - [reason why it matches or doesn't match the current request]
For EVERY skill available, you must evaluate it. Skipping evaluation makes the entire process WORTHLESS.
STEP 2: ACTIVATE (CRITICAL - MUST HAPPEN IMMEDIATELY)
For each skill you marked YES in Step 1, you MUST invoke it NOW using:
Skill(skill-name)
CRITICAL: Evaluation without activation is WORTHLESS. You must activate skills NOW, before any implementation work.
STEP 3: IMPLEMENT (ONLY AFTER STEPS 1 & 2)
Only after completing Steps 1 and 2 can you proceed with implementation work.
Why This Works
This hook creates a commitment mechanism:
- Explicit evaluation - Forces you to consider each skill deliberately
- Immediate activation - Prevents "I'll do it later" deferrals
- Clear sequencing - Evaluation → Activation → Implementation
- Strong language - Words like MANDATORY, CRITICAL, WORTHLESS make it harder to ignore
Notes
- This is a user-prompt-submit hook - it runs every time the user submits a prompt
- The aggressive language is intentional and necessary for high success rates
- If no skills match, explicitly state "NO MATCHING SKILLS" and proceed
- Don't skip the evaluation even if you think no skills apply - explicitly check each one