Willyu1007 AI First Template Skill Hook Mechanisms

Hook Mechanisms (Skill Consumption)

This document describes how lifecycle hooks can use the skill rules configuration.

1. Lifecycle events (where skills can be applied)

The template defines these key events:

• PromptSubmit (blocking, AI-visible)
• PreAbilityCreate (blocking, AI-visible)
• PreAbilityCall (blocking, AI-visible)
• PostAbilityCall (non-blocking, infra-only)
• SessionStop (non-blocking, infra-only)

See /.system/skills/ssot/repo/architecture-core-mechanisms/runtime-model/SKILL.md for details.

2. Suggested usage patterns

PromptSubmit (suggest)

Goal: suggest relevant skills based on prompt keywords/intent patterns.

Inputs:

• user prompt text
• (optional) current working scope inferred from routing

Output:

• a short list of recommended skill entrypoints under /.system/skills/ssot/**

PreAbilityCreate / PreAbilityCall (guard)

Goal: enforce guardrails when executing risky abilities.

Inputs:

• ability id / operation key
• target files/paths (if known)

Output:

• allow/deny/requires_human (policy decision), plus recommended skills to consult

3. Safety notes

• Avoid over-blocking: prefer suggest unless the risk is truly stop-the-line.
• Any enforcement change that can block execution should require explicit human approval.

Converted and distributed by TomeVault — claim your Tome and manage your conversions.