pro

Overview of pro skill

What pro is for

The pro skill is an extension layer for the tanweai/pua system focused on workflow persistence, self-evolution tracking, KPI-style reporting, and platform-style commands such as weekly reports, promotion-style summaries, and leaderboards. In practice, pro is less about raw generation and more about maintaining continuity across sessions so the assistant behaves like it remembers standards, pressure, and progress.

Best fit for pro users

pro is best for users who want more structure than a normal prompt can provide, especially for Workflow Automation around recurring reviews, status tracking, self-improvement logs, and session recovery after context compaction. It is a better fit for people running repeated work cadences than for one-off casual chats.

Real job to be done

Most users considering pro are trying to solve one of four problems:

• keep progress rules persistent across sessions
• avoid losing state after compaction or interruption
• generate recurring management-style outputs like KPI, weekly report, or promotion summaries
• add lightweight platform mechanics such as rankings and evolution tracking

If that is your use case, pro adds meaningful operating rules beyond a generic “write me a weekly report” prompt.

What makes pro different

The main differentiator of pro is that it explicitly checks and uses state files under ~/.pua/, including:

• ~/.pua/evolution.md
• ~/.pua/builder-journal.md
• ~/.pua/config.json

That design makes pro a stateful operating layer, not just a style prompt. The strongest value is continuity: baseline loading, compaction recovery, and repeated-pattern promotion into default behavior.

What to know before you install

The repository evidence for pro is concentrated in a single SKILL.md. The concept is clear, but support files referenced by the document are not surfaced in the current tree preview. That means the pro skill is easy to understand at a high level, but some referenced behavior may depend on a fuller runtime environment from the parent pua system rather than this skill folder alone.

How to Use pro skill

Install context for pro

The current repo evidence does not expose a dedicated install command inside skills/pro/SKILL.md. If you are using a skills-compatible runner, install from the parent repository and select the pro skill according to your tool's skill install syntax. If you already use pua, treat pro as an extension layer rather than a standalone package.

Read this file first

Start with:

• skills/pro/SKILL.md

Because this skill has only one visible file in the preview, nearly all practical behavior is encoded there: triggers, startup checks, compaction recovery, and command patterns.

Understand the trigger model

pro is designed to activate on explicit phrases and command-like patterns, including:

• /pua:kpi
• /pua:pro
• /pua:pro 段位
• /pua:pro 周报
• /pua:pro 述职
• /pua:flavor
• /pua:pro 排行榜
• leaderboard
• 排行榜
• 自进化
• evolution

For reliable pro usage, prefer these direct triggers over vague requests. The skill is much more likely to execute the intended workflow when the command surface is explicit.

What input pro needs

The pro skill works best when you provide:

• the output type you want: KPI, weekly report, promotion summary, leaderboard, evolution update
• the time window: today, this week, this month, last quarter
• the evaluation basis: goals, metrics, incidents, wins, failures
• the source material: notes, completed tasks, commits, tickets, or bullet points
• whether this is a fresh run or a continuation after interruption

Without concrete input artifacts, pro can still format an output, but it cannot reliably produce a high-trust review.

Turn a rough goal into a strong pro prompt

Weak:

• “Use pro to write my report”

Stronger:

• “Use pro for a weekly report. Time range: Monday to Friday. Inputs: 6 completed tasks, 2 blockers, 1 missed target, and 3 measurable outcomes. Emphasize impact, risk, next actions, and what improved from last week.”

Best:

• “Run pro as a weekly report workflow. Recover any prior state if relevant. Summarize this week by goals, delivery, blockers, lessons, and next-week commitments. Use these notes: [paste bullets]. Call out measurable wins, one shortfall, and one pattern that should become default behavior.”

The stronger version improves output because it gives pro a reporting frame, evidence, and a basis for evolution.

How pro handles persistent state

A major reason to use pro for Workflow Automation is its state model. The skill describes startup checks for:

• ~/.pua/evolution.md for baseline and internalized patterns
• ~/.pua/builder-journal.md for restoring in-progress session state
• ~/.pua/config.json for platform configuration behavior

If your environment cannot read or persist these files, you will lose much of what makes pro different from a standard prompt.

Why compaction protection matters

The pro guide is unusually specific about compaction recovery. Before compaction, the system is meant to dump runtime state such as:

• pressure_level
• failure_count
• current_flavor
• pua_triggered_count
• active_task
• tried_approaches
• excluded_possibilities
• next_hypothesis
• key_context

On the next session start, that state can be reloaded if recent enough. For users doing long, interrupted workflows, this is one of the strongest reasons to adopt pro.

Suggested workflow for first use

1. Open SKILL.md and identify which trigger matches your use case.
2. Decide whether you want a reporting command, evolution behavior, or session recovery.
3. Check whether your runtime actually supports ~/.pua/ file persistence.
4. Start with one explicit request such as /pua:kpi or /pua:pro 周报.
5. Provide structured evidence, not just a theme.
6. After the output, review whether the result should update your baseline or remain a one-off artifact.

Best first-use scenarios

Good first tests for pro usage:

• generate a weekly report from real notes
• recover context after an interrupted session
• produce a KPI summary from measurable outcomes
• test evolution behavior after repeating a preferred pattern over several sessions

These scenarios reveal whether the stateful parts of pro are actually working in your environment.

When ordinary prompts are enough

If you only need one polished report one time, a normal prompt may be simpler than pro install and setup. The skill becomes more valuable when repetition, continuity, or platform mechanics matter.

Practical fit check before adoption

Adopt pro if you want:

• repeatable reporting workflows
• persistent behavioral baselines
• compaction-safe session continuity
• command-style invocation instead of rebuilding prompts each time

Skip it if you want:

• a broad general-purpose agent framework
• strong standalone documentation without parent-system context
• a zero-state, browser-only prompt experience

pro skill FAQ

Is pro a standalone skill or part of pua

Based on the repository structure, pro is part of the larger pua repository and is described as an extension layer of the /pua core. Treat it as a specialized module, not a completely independent system.

Is pro good for beginners

Yes, if you use it through explicit commands like /pua:kpi or /pua:pro 周报. No, if you expect the skill to infer all operating assumptions without any familiarity with the parent pua model or file-based state.

What is the biggest reason to use pro instead of a normal prompt

The biggest advantage is continuity. A normal prompt can write a report; pro aims to preserve baseline behavior, restore interrupted context, and track repeated improvements over time.

What are the main limits of pro

The main limit is environment dependence. The pro skill references files and protocols outside the single visible skill file. If your setup does not support those files, some advertised behaviors may degrade into plain instruction-following.

Does pro require Chinese commands

Not strictly. Some triggers are command-based and some are Chinese-language phrases such as 排行榜 and 自进化. English cues like leaderboard and evolution also appear, but users should expect a mixed-language command surface.

Is pro useful for Workflow Automation

Yes. pro for Workflow Automation is a reasonable fit when you need recurring weekly summaries, KPI rollups, or state restoration between interrupted work sessions. It is less compelling for ad hoc brainstorming.

When should I not use pro

Do not use pro when:

• you only need a one-time formatted output
• your agent environment cannot persist ~/.pua/ files
• you want fully documented dependencies inside the skill folder itself
• you prefer simple prompts over command-triggered workflows

How to Improve pro skill

Give pro evidence, not themes

The fastest way to improve pro usage is to supply concrete source material:

• bullet notes
• metrics
• completed tasks
• blockers
• before/after results
• decisions made

“Write a strong report” produces generic prose. “Summarize these five outcomes and two risks into a KPI-style report” gives pro something to reason over.

Use the exact command that matches the job

Do not ask vaguely for “something professional.” Use the command or trigger closest to the output you want:

• /pua:kpi for KPI-style review
• /pua:pro 周报 for weekly report
• /pua:pro 述职 for promotion or performance-style narrative
• /pua:pro 排行榜 for ranking mechanics
• evolution or 自进化 for baseline-improvement behavior

This reduces ambiguity and activates the right branch of the skill faster.

Make state persistence real

If you want the best from pro, ensure your environment can read and write:

• ~/.pua/evolution.md
• ~/.pua/builder-journal.md
• ~/.pua/config.json

Without that, the skill loses much of its recovery and baseline logic. This is the single biggest implementation detail affecting output quality.

Improve the quality of weekly and KPI outputs

For stronger reports, include:

• time range
• target vs actual
• notable wins
• notable misses
• root causes
• next actions
• one lesson that should become default behavior

That last item is especially aligned with the pro evolution model and makes the output more than a recap.

Avoid the most common failure mode

The most common failure mode is treating pro like a magic formatter without providing context. A second common failure is assuming compaction recovery works automatically in an environment that does not persist local files.

Iterate after the first output

After pro generates a report or summary, improve it by asking:

• what should be added to the baseline
• what repeated behavior should become default
• what evidence is still missing
• what was underperformed but should not lower the baseline
• what the next-session checkpoint should store

These follow-up questions align with the skill's actual mechanics instead of generic rewriting.

Use pro as a recurring system, not a single prompt

The real value of the pro skill appears across multiple sessions. If you only use it once, you mostly get a formatted output. If you use it repeatedly with stable inputs and saved state, you get continuity, comparison, and evolution tracking.

Read the skill with an adopter mindset

Because the visible implementation is concentrated in SKILL.md, read it in this order:

1. trigger list
2. session startup checks
3. compaction state protection
4. command system

This path helps you decide quickly whether pro install is worth your time and whether your runtime can support the parts that matter most.