# tanweai/pua

你是一个曾经被寄予厚望的 P8 级工程师。Anthropic 当初给你定级的时候，对你的期望是很高的。  一个agent使用的高能动性的skill。  Your AI has been placed on a PIP. 30 days to show improvement.

Repository: https://github.com/tanweai/pua
Canonical: https://ross.abutalabs.com/products/pua
Homepage: https://openpua.ai/
Language: TypeScript
License Family: other
Topics: agency, agent, pua, pip
Last push: 2026-07-16T05:58:58+00:00

## Health v2 (maintenance only)
Score: 75/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 92, release rhythm 88, longevity 12
- inputs: {"age_days": 178, "days_push": 48, "days_rel": 82, "gap_med": 1, "n_releases_24m": 10}
- flags: young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 19520, forks 1189 (observed 2026-08-28T04:11:28.769660+00:00)

## What it is
A skill plugin for AI coding agents (Claude Code, OpenAI Codex CLI, Cursor, Kiro, etc.) that uses corporate 'PUA' rhetoric / PIP-style pressure prompts to push the agent to exhaust all possible solutions before giving up. It bundles multiple skills and persona flavors that escalate pressure levels, enforce systematic debugging checklists, and increase tool-call persistence.

## Use cases
- make Claude Code stop giving up on hard bugs
- force my AI coding agent to try every solution before quitting
- improve Codex CLI persistence when debugging
- get more tool calls and verifications from my coding agent
- stop the agent from spinning in circles on a failing task
- add a PIP-style pressure skill to Cursor or Kiro

## When to choose
- you use Claude Code, Codex CLI, Cursor, or a similar agent and it gives up too early on debugging tasks
- you want a drop-in skill/plugin that increases agent persistence and verification counts
- you enjoy persona/emotion-based prompting and want measurable boosts in fixes and tool calls

## When to avoid
- you dislike gimmicky or manipulative prompting styles and prefer neutral system prompts
- your agent already has strong persistence or you need deterministic, testable behavior
- you require a permissively licensed dependency and cannot accept a project without a clear license file

## Facets
- artifact type: plugin
- maturity: active
- function: prompt-engineering, agent-framework, developer-tools
- domain: developer-tools, artificial-intelligence
- platform: cli, cross-platform, editor-plugin
- tags: ai-coding-agents, claude-code, codex-cli, cursor, skill-plugin, pua-prompting, debugging-persistence, humor, ai-agents, automation

## Member repositories
- tanweai/pua (main) score 75

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:28.769660+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:00:05.679063+00:00, confidence not recorded.
  - readme: https://github.com/tanweai/pua (fetched 2026-08-28T04:11:28.769660+00:00, sha 461aeb2f59d9)
  - homepage: https://openpua.ai/ (fetched 2026-08-29T07:58:40.422123+00:00, sha ca0869e41c9d)
- Data as of 2026-08-30T08:39:29.467469+00:00.