jianshuo/claude-skills / wjs-x-improving-content
wjs-x-improving-content
VerifiedUse when 王建硕 wants to systematically improve his X (Twitter) content by iterating on the content-generation prompt (prompts/x/prompt.md, used by the every-6h tweet Action) and finding which prompt version produces the highest-reach tweets. Each prompt edit is a git-SHA-versioned, numbered experiment with a hypothesis; tweets are attributed to the version live at post time and judged on median impressions per tweet. Also mines per-tweet impression data for content-feature signals (angle / length / topic) that feed the next prompt edit. North-star = impressions per tweet. Triggers — "改 X 的 prompt", "X 内容改进", "哪版 prompt 最好", "什么内容 impression 高", "improve my tweets", "A/B test the X prompt", "/wjs-x-improving-content".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 6.2 KB |
| inbox/.gitignore | other | 12 B |
| scripts/_common.pyexecutable | script | 1.5 KB |
| scripts/analyze-content.pyexecutable | script | 4.1 KB |
| scripts/evaluate.pyexecutable | script | 3.1 KB |
| scripts/ingest-tweets.pyexecutable | script | 4.3 KB |
| scripts/ledger.pyexecutable | script | 2.9 KB |
| scripts/scoreboard.pyexecutable | script | 3.7 KB |
| state/.gitignore | other | 6 B |
| state/SCOREBOARD.md | markdown | 1.1 KB |
| state/tweets.jsonl | other | 42.1 KB |
| state/versions.jsonl | other | 523 B |
scripts/ingest-tweets.py
View on GitHub#!/usr/bin/env python3
"""Join the X Analytics Content CSV with the bot's post history into state/tweets.jsonl.
Each row of the Content CSV is one tweet (Post id = tweet_id). We:
- parse per-tweet impressions + engagement counts
- join to history.jsonl on tweet_id → adds slug, angle, source=bot (else manual)
- derive prompt_sha by date: the prompts/x/prompt.md git commit live at post time
(tweets posted before the prompt file existed get prompt_sha=null = "pre-prompt")
- compute char_len (CJK proxy) + age_days + mature flag
Usage: python3 ingest-tweets.py /path/to/content.csv
Upserts by tweet_id. Re-running with a longer export is safe.
"""
import argparse, csv, subprocess, sys
from datetime import datetime, date
from pathlib import Path
sys.path.insert(0, str(Path(__file__).resolve().parent))
from _common import (TWEETS, HISTORY, REPO, PROMPT_FILE, MATURITY_DAYS, # noqa: E402
read_jsonl, write_jsonl, cjk_len)
def parse_date_cell(v):
v = v.strip().strip('"')
for fmt in ("%a, %b %d, %Y", "%Y-%m-%d", "%b %d, %Y", "%m/%d/%Y", "%a %b %d, %Y"):
try:
return datetime.strptime(v, fmt).date().isoformat()
except ValueError:
continue
return None
def to_int(v):
try:
return int(round(float(str(v).strip().replace(",", "") or 0)))
except ValueError:
return None
def prompt_commits():
"""[(iso_date, short_sha)] for prompts/x/prompt.md, oldest→newest. Empty if none."""
try:
out = subprocess.run(
["git", "-C", str(REPO), "log", "--reverse", "--date=short",
"--format=%ad %h", "--", PROMPT_FILE],
capture_output=True, text=True, check=True).stdout.strip()
except Exception:
return []
rows = []
for line in out.splitlines():
if line.strip():
d, sha = line.split()
rows.append((d, sha))
return rows
def sha_for_date(commits, d_iso):
"""latest prompt commit with date <= tweet date; None if tweet predates the prompt file."""
chosen = None
for cdate, sha in commits:
if cdate <= d_iso:
chosen = sha
else:
break
return chosen
def main():
ap = argparse.ArgumentParser()
ap.add_argument("csv_path")
args = ap.parse_args()
hist = {h["tweet_id"]: h for h in read_jsonl(HISTORY) if h.get("tweet_id")}
commits = prompt_commits()
today = date.today()
existing = {t["tweet_id"]: t for t in read_jsonl(TWEETS)}
rows = list(csv.DictReader(open(args.csv_path, newline="", encoding="utf-8-sig")))
n = 0
for r in rows:
tid = (r.get("Post id") or "").strip()
d = parse_date_cell(r.get("Date", ""))
if not tid or not d:
continue
text = r.get("Post text", "") or ""
h = hist.get(tid, {})
age = (today - date.fromisoformat(d)).days
rec = {
"tweet_id": tid,
"date": d,
"impressions": to_int(r.get("Impressions")),
"engagements": to_int(r.get("Engagements")),
"likes": to_int(r.get("Likes")),
"replies": to_int(r.get("Replies")),
"reposts": to_int(r.get("Reposts")),
"new_follows": to_int(r.get("New follows")),
"char_len": cjk_len(text),
"text": text,
"slug": h.get("slug"),
"angle": h.get("angle"),
"source": "bot" if tid in hist else "manual",
"prompt_sha": sha_for_date(commits, d),
"age_days": age,
"mature": age >= MATURITY_DAYS,
}
existing[tid] = rec
n += 1
write_jsonl(TWEETS, [existing[k] for k in sorted(existing, key=lambda k: existing[k]["date"])])
mature = sum(1 for t in existing.values() if t["mature"])
bot = sum(1 for t in existing.values() if t["source"] == "bot")
print(f"Upserted {n} tweet(s). Total {len(existing)} ({bot} bot / {len(existing)-bot} manual), "
f"{mature} mature (≥{MATURITY_DAYS}d).")
if not commits:
print("⚠ no git history for prompts/x/prompt.md — prompt_sha=null for all. "
"Version comparison starts once the prompt file has commits.")
else:
print(f"prompt versions seen: {len(commits)} ({', '.join(s for _, s in commits)})")
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |