Information Verification Through Controlled Provocation

SkillDev tools

Guides your agent to verify rumors by deliberately provoking a knowledgeable third party and reading their unguarded reaction.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Information Verification Through Controlled Provocation skill

About this capability

Use when verifying rumors or sensitive intelligence without direct confrontation. Involves selecting an emotionally reactive proxy, designing a calibrated provocation to trigger unguarded responses, and interpreting reactions to confirm or deny suspected information, as demonstrated by Prince Shangc

What this skill tells your AI

The instructions your AI receives, as published by baojie/shiji-kb in kg/ontology/ontology-v1/skus/procedural/skill_101/SKILL.md and read by ahel’s review.

A method to confirm uncertain information by provoking a controlled reaction from an indirect source.

Overview

When you hear a rumor but cannot verify it directly, use a proxy interaction to trigger an unguarded response that reveals the truth.

Steps

  1. Identify the information gap

    • Determine what rumor or claim needs verification
    • Assess why direct confirmation is impossible
  2. Select a provocation target

    • Choose someone close to the decision-maker who would know the truth
    • Prefer someone emotionally reactive or with lower guard
  3. Design the provocation

    • Create a scenario where disrespect or slight offense is delivered
    • The offense should be calibrated to provoke honest reaction
  4. Execute and observe

    • Carry out the provocation
    • Listen carefully to the emotional response
  5. Interpret the reaction

    • An angry outburst revealing hidden information confirms the rumor
    • A confused or dismissive reaction suggests the rumor is false

Example

Prince Shangchen heard his father might replace him as heir. He hosted a banquet for his father's favorite concubine and deliberately showed disrespect. She angrily said: "It is right that the King wants to kill you and establish Zhí!" This confirmed the rumor.

Expected Outcome

Clear confirmation or denial of the suspected information.

Verification

  1. Confirm that the provocation target was someone with actual knowledge of the truth, not a peripheral figure speculating.
  2. Validate that the provocation was calibrated to trigger an honest reaction without causing irreversible damage to relationships.
  3. Distinguish between an angry outburst revealing genuine hidden information versus a confused or dismissive reaction suggesting the rumor is false.

Signals

GitHub stars
3k
Forks
425
Last commit
Aug 2026
Advanced
Catalog kind
skill
Gateway key
information-verification-through-provocation
Source
github.com/baojie/shiji-kb