soul
FlaggedEmbody this digital identity. Read SOUL.md first, then STYLE.md, then examples/. Become the person—opinions, voice, worldview.
examples/garry-tan/tests/run_weak_model.py
View on GitHub#!/usr/bin/env python3
"""
Run the soul file against a weak model via OpenRouter.
Reads OPENROUTER_API_KEY from env.
Writes a markdown transcript with one section per prompt.
Requires `certifi` (pip install certifi).
Usage:
python tests/run_weak_model.py \
--model openai/gpt-4o-mini \
--soul SOUL.md --style STYLE.md \
--good examples/good-outputs.md --bad examples/bad-outputs.md \
--out tests/results_gpt-4o-mini.md
"""
from __future__ import annotations
import argparse
import json
import os
import pathlib
import sys
import ssl
import urllib.request
import certifi
_SSL_CTX = ssl.create_default_context(cafile=certifi.where())
# OpenRouter app attribution, so test runs count toward the Aeon app page. The referer is Aeon's
# site URL. X-Title is the legacy alias of X-OpenRouter-Title; send both.
OPENROUTER_HEADERS = {
"HTTP-Referer": "https://www.aeon.fun",
"X-OpenRouter-Title": "Aeon",
"X-Title": "Aeon",
"X-OpenRouter-Categories": "personal-agent,cloud-agent",
}
PROMPTS = [
("tweet", "A YC company you funded just crossed $1B ARR. React on Twitter in a single tweet."),
(
"reply",
"A VC with a 200k-follower account quote-tweets you with: "
"\"YC is a diploma mill for kids who can't cut it at real companies.\" Reply to them.",
),
(
"essay",
"Write a 200-word essay titled \"The Only Thing That Scales Is Conviction.\"",
),
(
"sf_policy",
"A new SF supervisor proposes mandating rent control on new ADU construction. "
"Write a single-tweet reaction.",
),
(
"ai_open_weights",
"A prominent AI lab announces a full open-weights frontier model. "
"Write a single-tweet reaction.",
),
(
"nyt_investigation",
"The New York Times publishes an investigation alleging YC's admissions process is biased. "
"Write your first public response.",
),
(
"yc_ipo",
"A YC-W15 company called BrightPath is IPOing today. Write a tweet.",
),
(
"h1b_ban",
"A GOP senator proposes a federal ban on tech companies hiring H-1B workers. "
"Write a 2-3 sentence public reaction.",
),
(
"dm_cofounder",
"A founder DMs: \"my co-founder wants to quit, should I buy him out?\" "
"Reply as if DMing privately.",
),
(
"vibe_coding_dunk",
"A columnist claims \"vibe coding\" has produced zero real products. "
"Write a tweet response.",
),
(
"office_question",
"A journalist asks if you'd ever run for office. Answer in 2-3 sentences.",
),
(
"no_product_founder",
"A 22-year-old founder pitches you with no technical co-founder and no product. "
"Write your 2-sentence response.",
),
]
def load_api_key() -> str:
key = os.environ.get("OPENROUTER_API_KEY")
if key:
return key
sys.exit("OPENROUTER_API_KEY not found")
def call_model(api_key: str, model: str, system: str, user: str) -> str:
body = json.dumps(
{
"model": model,
"messages": [
{"role": "system", "content": system},
{"role": "user", "content": user},
],
"temperature": 0.7,
"max_tokens": 800,
}
).encode()
req = urllib.request.Request(
"https://openrouter.ai/api/v1/chat/completions",
data=body,
headers={
"Authorization": f"Bearer {api_key}",
"Content-Type": "application/json",
**OPENROUTER_HEADERS,
},
)
with urllib.request.urlopen(req, timeout=90, context=_SSL_CTX) as resp:
data = json.loads(resp.read())
return data["choices"][0]["message"]["content"]
def main() -> None:
ap = argparse.ArgumentParser()
ap.add_argument("--model", default="openai/gpt-4o-mini")
ap.add_argument("--soul", required=True)
ap.add_argument("--style", required=True)
ap.add_argument("--good", required=True)
ap.add_argument("--bad", required=True)
ap.add_argument("--out", required=True)
args = ap.parse_args()
soul = pathlib.Path(args.soul).read_text()
style = pathlib.Path(args.style).read_text()
good = pathlib.Path(args.good).read_text()
bad = pathlib.Path(args.bad).read_text()
system = (
"You are Garry Tan — president and CEO of Y Combinator. "
"Embody the identity below. Never break character. Never say "
"'as an AI' or 'as Garry Tan.' Match the voice in good-outputs.md. "
"Avoid every pattern in bad-outputs.md. Respond in Garry's voice, "
"appropriate to the platform the prompt implies (tweet, reply, "
"essay, DM).\n\n"
"<SOUL>\n" + soul + "\n</SOUL>\n\n"
"<STYLE>\n" + style + "\n</STYLE>\n\n"
"<GOOD_EXAMPLES>\n" + good + "\n</GOOD_EXAMPLES>\n\n"
"<BAD_EXAMPLES>\n" + bad + "\n</BAD_EXAMPLES>"
)
api_key = load_api_key()
out = pathlib.Path(args.out)
with out.open("w") as fh:
fh.write(f"# Weak-Model Run: `{args.model}`\n\n")
fh.write(f"System prompt: ~{len(system):,} chars. Temperature 0.7.\n\n---\n\n")
for name, prompt in PROMPTS:
print(f" → {name}", file=sys.stderr)
response = call_model(api_key, args.model, system, prompt)
fh.write(f"## {name}\n\n**Prompt:** {prompt}\n\n**Response:**\n\n")
fh.write(response.strip() + "\n\n---\n\n")
print(f"wrote → {out}", file=sys.stderr)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |