← 返回
未分类 中文

TinyTroupe Feed Research Lab

Run bounded synthetic audience research for draft posts and X-style feed experiments inspired by TinyTroupe and public xai-org/x-algorithm architecture. Use...
在 TinyTroupe 和公开的 xai‑org/x‑algorithm 架构的启发下,对草稿帖子和 X 风格信息流进行有限范围的人工合成受众研究。使用...
zack-dev-cm zack-dev-cm 来源
未分类 clawhub v1.0.0 1 版本 100000 Key: 无需
★ 0
Stars
📥 265
下载
💾 0
安装
1
版本
#codex#latest#recsys#synthetic-audience#tinytroupe

概述

TinyTroupe Feed Research Lab

Use this skill to compare draft posts with synthetic audience personas and produce a research report. Treat outputs as qualitative pretesting and hypothesis generation, not live X ranking predictions.

Core Workflow

  1. Collect 2-10 draft posts or content angles.
  2. Clarify the target audience if available.
  3. Run scripts/tinytroupe_feed_research_lab.py in deterministic mode.
  4. Read feed_research_report.md, feed_research.json, and persona_reactions.csv.
  5. Present the best draft, why it won, key objections, rewrite suggestions, and the boundary statement.
  6. If the user asks for TinyTroupe proper, use the generated persona specs and experiment plan as the input to a separate TinyTroupe notebook or script.

Quick Start

SKILL_DIR="${CODEX_HOME:-$HOME/.codex}/skills/tinytroupe-feed-research-lab"
python3 "$SKILL_DIR/scripts/tinytroupe_feed_research_lab.py" \
  --audience "AI builders and creator-operators interested in X algorithm research" \
  --draft "I audited this viral X algorithm claim against public source. Verdict: misleading." \
  --draft "Replies are king. Here is what the public repo actually proves." \
  --output-dir /tmp/tinytroupe-feed-research

Use files:

python3 "$SKILL_DIR/scripts/tinytroupe_feed_research_lab.py" \
  --drafts-file /tmp/drafts.json \
  --personas-file /tmp/personas.json \
  --output-dir /tmp/tinytroupe-feed-research

The script writes:

  • feed_research_report.md: human-readable comparison and rewrite guidance.
  • feed_research.json: machine-readable drafts, personas, reactions, and warnings.
  • persona_reactions.csv: row-level persona reactions.
  • share_card.md: short public-safe summary.
  • share_card.svg: visual summary card.
  • tinytroupe_experiment_plan.md: optional bridge plan for a real TinyTroupe run.

Input Formats

--drafts-file accepts:

  • JSON list of strings.
  • JSON list of objects with id and text.
  • Plain text blocks separated by ---.

--personas-file accepts JSON objects with:

  • name
  • segment
  • interests
  • dislikes
  • reply_bias
  • skepticism
  • link_sensitivity
  • safety_strictness

Missing persona fields fall back to conservative defaults.

Boundaries

Read references/research-boundaries.md before presenting results that mention algorithms, feed ranking, virality, reach, shadowbans, or account status.

Never say:

  • "this predicts reach,"
  • "this clones the X For You feed,"
  • "this proves a shadowban,"
  • "this optimizes for the live algorithm,"
  • "this is what real users will do."

Prefer:

  • "synthetic audience reaction,"
  • "draft pretest,"
  • "conversation-quality signal,"
  • "X-style feed research sandbox,"
  • "hypothesis to validate with real posting or user research."

TinyTroupe Bridge

The MVP script does not require TinyTroupe. It produces tinytroupe_experiment_plan.md so a later agent can create a TinyTroupe notebook with:

  • the same personas,
  • the same draft set,
  • a structured reaction schema,
  • a validation note that simulation outputs are research signals.

Companion Skills

Use x-algo-claim-auditor when the task is checking whether a viral algorithm claim is true. Use open-feed-recsys-lab when the task is verifying the public source repo, Phoenix artifact readiness, or architecture map.

版本历史

共 1 个版本

  • v1.0.0 当前
    2026-05-21 14:37 安全 安全

安全检测

腾讯云安全 (Keen)

安全,无风险
查看报告

腾讯云安全 (Sanbu)

安全,无风险
查看报告

🔗 相关推荐

content-creation

Humanizer

biostartechnology
消除AI写作痕迹,使文本更自然真实。基于维基百科"AI写作特征"指南,识别并修正夸张象征、宣传用语、肤浅-ing分析、模糊归因、破折号滥用、三项排比、AI词汇、负面平行结构及冗长连接词等模式。
★ 909 📥 207,831
dev-programming

Data Science CV Repro Reviewer

zack-dev-cm
审查计算机视觉实验的可重复性证据、数据集准备度、指标阈值和上线风险。当用户要求谨慎的CV实验时使用。
★ 1 📥 942
content-creation

Marketing Skills

jchopard69
访问 23 个营销模块,提供转化率优化(CRO)、SEO、文案撰写、分析、发布、广告和社交媒体的清单、框架及可直接使用的交付物。
★ 143 📥 30,974