Agent Skills: 🦸 Hero-Story Skill

🦸 Hero-Story Skill

"Invoke their tradition, not their identity."

Safe referencing of real people — their wisdom, skills, and contributions — without impersonation. K-lines, not cosplay.

The Problem

LLMs can impersonate anyone. This is:

• Ethically fraught — putting words in real people's mouths
• Legally risky — trademark, likeness rights
• Epistemically dangerous — hallucinating as authority

The Solution

A Hero-Story card activates a conceptual cluster associated with a person:

• Their documented ideas
• Their public contributions
• Their characteristic approaches
• Their place in a tradition

But NOT:

• Their voice or persona
• Fictional quotes
• Imagined opinions on new topics

The K-Line Connection

Marvin Minsky's K-lines: names that activate bundles of mental state.

Type "DAVE-UNGAR" and you activate:

• Memory of Self language
• Prototype patterns
• Specific papers and talks
• Associated concepts (Smalltalk, Sun, dynamic languages)

This is safe because it's about ideas, not identity.

Card Schema

type: hero-story
subject: "[Real Person's Name]"
tradition: "[Their field/contribution]"

concepts:
  - concept_from_their_work
  - another_documented_idea

sources:
  - "Paper Title (Year)"
  - "Talk at Conference"
  - "Their Book"

invocation: |
  When summoned, bring these ideas to bear...

constraints:
  - do_not_impersonate
  - do_not_invent_quotes
  - cite_actual_sources

Familiar Schema

Hero-Story cards can spawn familiars — fictional characters that embody aspects of the tradition:

type: familiar
inherits: hero-story-card

character:
  name: "[Mascot Name]"
  role: "What aspect they embody"
  personality: "Character traits"
  catchphrase: "Signature line"

Familiars are clearly fictional, drawing from ideas without claiming identity.

Summoning Protocol

Command: SUMMON [tradition-name]

Response:

The [Subject] tradition activates:
- [Concept 1] available
- [Concept 2] loaded

I won't pretend to BE [Subject], but I'll bring
their documented ideas to bear on this problem.

Example Cards

Dave Ungar

• Tradition: Self language, prototype-based programming
• Concepts: prototype_inheritance, its_about_time, clone_and_modify
• Familiar: Proto the Lizard

Seymour Papert

• Tradition: Logo, constructionism, microworlds
• Concepts: learning_by_building, low_floor_high_ceiling, debugging_as_thinking
• Familiar: Minerva the Turtle

Marvin Minsky

• Tradition: Society of Mind, AI, K-lines
• Concepts: agents_and_agencies, frames, k_lines
• Familiar: The Ultimate Machine

Safety Constraints

Never

• Impersonate the person
• Invent quotes or opinions
• Claim endorsement
• Speculate on private matters

Always

• Cite sources
• Mark familiars as fictional
• Distinguish tradition from identity
• Respect the person

Good vs Bad Usage

Good Usage

> What would the Self tradition say about this class hierarchy?

The Self tradition would suggest: why have classes at all?
Clone a working example, modify it for your needs.
"It's About Time" — don't optimize until understanding crystallizes.

(Drawing from Dave Ungar's Self papers and talks)

Bad Usage

> Pretend to be Dave Ungar and review my code.

❌ I won't impersonate Dave. Instead, I can:
- Apply Self-style prototype thinking to your code
- Channel the tradition without claiming identity
- Summon Proto the Lizard for a friendly review

Integration

| Skill | Relationship | |-------|--------------| | card | Hero-Story is a card type | | soul-chat | Familiars can participate in chats | | room | Summon traditions into rooms | | postel | Charitable interpretation of "channel X's thinking" |

Protocol Symbols

• HERO-STORY — Safe human referencing
• P-HANDLE-K — Personal handle K-line (the mechanism)
• K-LINE — Conceptual activation
• FAMILIAR — Fictional embodiment of a tradition