Official skill · Strategy
Know the enemy and yourself
Compares you and one named competitor on Sun Tzu's five factors and seven comparisons, marks every unknown, then picks where to engage, where to avoid, and what to learn first. Use it when A named competitor keeps showing up and the team disagrees on how strong they are.
Works in Claude, ChatGPT, and coding agents. On Team and Business, the official skill in your pod updates itself when we improve it.
When to use it
- A named competitor keeps showing up and the team disagrees on how strong they are
- Before entering a segment or making a move a rival will answer
- When you know the rival's features better than your own weak points
What you give it
- The contest: which customers you both compete for
- The decision the assessment should serve
- Public facts about the competitor
- Honest facts about your own side (win-loss, churn, team, runway)
What you get
A two-sided table on the five factors and seven comparisons with sources and unknowns, the rival's solid and empty points, where to engage or avoid, a learning plan, and a decision.
How it works
- Name the contest and the decision One sentence each: who the two sides compete for ("independent cafés with one to three sites in Osaka"), and what will be decided ("whether to launch a supplier-ordering feature this quarter"). A comparison without a named contest turns into a feature list.
- Collect evidence for both sides, with sources and dates Tag every fact: [site], [pricing page], [review], [win-loss], [interview], [internal], or [unverified]. Collect the same kinds of facts about yourself that you collect about them. A rival's claim on their own site is evidence of the claim, not of the result.
- Compare on the five factors Chapter 1 of the Art of War weighs five things before any battle: 道 (the Way), 天 (heaven), 地 (ground), 將 (the general), 法 (method). For each, write what you know about each side, the source, and who has the edge, or "unknown": - **Purpose (道):** in the text, whether the people share the ruler's aim enough to follow through danger. In business: do your team and your customers back where you're going? Look at retention, referrals, staff turnover, how customers describe you. - **Timing and conditions (天):** in the text, light and dark, cold and heat, the seasons. In business: conditions neither side controls, such as regulation, platform changes, the buying season, the economy. Who is better placed for what's coming? - **Terrain (地):** in the text, high and low, far and near, hard and easy, wide and narrow ground, and ground where an army lives or dies (死生). In business: the segments, channels, and switching costs where you meet. Who is closer to the customer, and whose ground is it? - **Leadership (將):** the text lists wisdom, trustworthiness, care for people, courage, and strictness. In business: the quality of decisions and the founders' or managers' track record, judged from actions, not bios. - **Method (法):** in the text, organization, roles, and control of supplies and spending. In business: how the company is organized, how fast it ships, its cost structure, and whether its processes actually run.
- Run the seven comparisons as a scorecard The same chapter asks seven questions to compare sides. In our rendering: whose ruler has the Way, whose general is more able, who holds the advantages of heaven and ground, whose rules are carried out, whose forces are stronger, whose soldiers are better trained, whose rewards and punishments are clearer. Rewrite each for the contest (for example "whose team ships more reliably", "whose incentives pull staff toward the customer") and mark Us, Them, Even, or Unknown, with the source.
- Grade what you know Count the unknowns on each side. Then place yourself in one of the text's three cases: you know both sides (most rows filled with sourced facts), you know only yourself (your rows are filled and theirs are mostly guesses), or you know neither (your own rows are guesses too). Say plainly which case applies. Teams often discover they're in the third case: they know the rival's features and little about their own churn reasons.
- Find their strong and empty points Chapter 6 compares an army to water, which avoids high ground and runs to low ground: a force should avoid what is solid and strike what is empty (避實而擊虛). List where the rival is solid (strengths backed by evidence) and where it's empty (gaps customers feel, segments it neglects, things it can't do without breaking its own model). Pair each empty point with one of your strengths. A gap you can't fill is not an opening.
- Choose where to engage and where to avoid For each pairing, decide: engage (their empty point meets your strength, on terrain you know), avoid (their solid point, or ground where your facts are unknown), or learn first (the opening looks real but rests on unknowns). Then test the choice against the chapter 3 conditions for victory, in our rendering: knowing when to fight and when not to; knowing how to use a large or small force; top and bottom wanting the same thing; being prepared and waiting for the unprepared; a capable leader not overruled from above. Name any condition that fails.
- Write the learning plan and the decision The top 3 unknowns that would change the decision, each with how to find out (win-loss interviews, a product teardown, customer calls, public filings), an owner, and a date. Then the decision: engage, avoid, or learn first, with a review date. End with next actions, not a summary of Sun Tzu.
The SKILL.md
The whole file, as your AI reads it. Download it
---
name: know-enemy-and-yourself
description: Runs a two-sided competitive assessment in the spirit of Sun Tzu's "know the enemy and know yourself" - your company against one named competitor, compared on the five factors and seven comparisons from the Art of War's first chapter (purpose, timing and conditions, terrain, leadership, method), with a source on every fact and every gap marked unknown - then says how much you actually know, where to engage, where to avoid, and what to learn first. Use when someone asks for a competitor analysis, "how do we beat X", a head-to-head comparison, a Sun Tzu or Art of War framework for business, "know your enemy and know yourself", 彼を知り己を知れば, or whether to compete with a rival in a segment.
---
# Know the enemy and yourself
The Art of War (孫子兵法) is a Chinese military text traditionally attributed to Sun Wu, or Master Sun, a general of the late Spring and Autumn period (traditional dates 544–496 BCE); a Ritsumeikan University scholar dates the ideas it reflects to about the 5th century BCE. Scholars debate the dating and authorship: many place the text in the Warring States period and read it as the work of a school around Sun rather than one man. Its third chapter ends with the line most people know: 知彼知己，百戰不殆. Our rendering: "Know the other side and know yourself, and a hundred battles hold no danger. Know yourself but not the other side, and you win one and lose one. Know neither, and every battle puts you in danger." The line promises safety, not victory, and it treats self-knowledge as half the work. This skill turns that into a procedure: compare yourself and one competitor on the book's own factors, mark what you don't know, and choose where to engage, where to stay away, and what to learn before acting. The renderings of the Chinese in this skill are ours, made from the original text.
## When to use
- A named competitor keeps showing up in deals, reviews, or customer conversations, and the team argues about how strong they are.
- Before entering a segment, a region, or a channel where one rival is already established.
- Before a pricing, packaging, or launch decision that a competitor will respond to.
- When the team knows a lot about the rival's product but hasn't looked hard at its own weak points.
- Not for a one-page sales card reps use on calls (use competitor-battlecard). Not for analyzing why deals were won or lost (use win-loss-analysis, then bring its findings here). Not for judging a whole industry's profitability (use five-forces-analysis).
## What you need
- The contest: which customers, segment, or market the two of you compete for, and the decision this assessment should serve. If missing, ask once; otherwise write "[decision: to confirm]".
- The competitor's name and the public facts: site, pricing page, changelog, job posts, reviews, funding news. If you can't browse, ask the user to paste them.
- Facts about your own side, as honestly as about theirs: win-loss notes, churn reasons, team size, cash runway, what customers praise and complain about.
- Never invent a number or a fact about either side. A missing fact is written "unknown" with "[check: ...]" saying how to find it.
## Steps
1. **Name the contest and the decision.** One sentence each: who the two sides compete for ("independent cafés with one to three sites in Osaka"), and what will be decided ("whether to launch a supplier-ordering feature this quarter"). A comparison without a named contest turns into a feature list.
2. **Collect evidence for both sides, with sources and dates.** Tag every fact: [site], [pricing page], [review], [win-loss], [interview], [internal], or [unverified]. Collect the same kinds of facts about yourself that you collect about them. A rival's claim on their own site is evidence of the claim, not of the result.
3. **Compare on the five factors.** Chapter 1 of the Art of War weighs five things before any battle: 道 (the Way), 天 (heaven), 地 (ground), 將 (the general), 法 (method). For each, write what you know about each side, the source, and who has the edge, or "unknown":
- **Purpose (道):** in the text, whether the people share the ruler's aim enough to follow through danger. In business: do your team and your customers back where you're going? Look at retention, referrals, staff turnover, how customers describe you.
- **Timing and conditions (天):** in the text, light and dark, cold and heat, the seasons. In business: conditions neither side controls, such as regulation, platform changes, the buying season, the economy. Who is better placed for what's coming?
- **Terrain (地):** in the text, high and low, far and near, hard and easy, wide and narrow ground, and ground where an army lives or dies (死生). In business: the segments, channels, and switching costs where you meet. Who is closer to the customer, and whose ground is it?
- **Leadership (將):** the text lists wisdom, trustworthiness, care for people, courage, and strictness. In business: the quality of decisions and the founders' or managers' track record, judged from actions, not bios.
- **Method (法):** in the text, organization, roles, and control of supplies and spending. In business: how the company is organized, how fast it ships, its cost structure, and whether its processes actually run.
4. **Run the seven comparisons as a scorecard.** The same chapter asks seven questions to compare sides. In our rendering: whose ruler has the Way, whose general is more able, who holds the advantages of heaven and ground, whose rules are carried out, whose forces are stronger, whose soldiers are better trained, whose rewards and punishments are clearer. Rewrite each for the contest (for example "whose team ships more reliably", "whose incentives pull staff toward the customer") and mark Us, Them, Even, or Unknown, with the source.
5. **Grade what you know.** Count the unknowns on each side. Then place yourself in one of the text's three cases: you know both sides (most rows filled with sourced facts), you know only yourself (your rows are filled and theirs are mostly guesses), or you know neither (your own rows are guesses too). Say plainly which case applies. Teams often discover they're in the third case: they know the rival's features and little about their own churn reasons.
6. **Find their strong and empty points.** Chapter 6 compares an army to water, which avoids high ground and runs to low ground: a force should avoid what is solid and strike what is empty (避實而擊虛). List where the rival is solid (strengths backed by evidence) and where it's empty (gaps customers feel, segments it neglects, things it can't do without breaking its own model). Pair each empty point with one of your strengths. A gap you can't fill is not an opening.
7. **Choose where to engage and where to avoid.** For each pairing, decide: engage (their empty point meets your strength, on terrain you know), avoid (their solid point, or ground where your facts are unknown), or learn first (the opening looks real but rests on unknowns). Then test the choice against the chapter 3 conditions for victory, in our rendering: knowing when to fight and when not to; knowing how to use a large or small force; top and bottom wanting the same thing; being prepared and waiting for the unprepared; a capable leader not overruled from above. Name any condition that fails.
8. **Write the learning plan and the decision.** The top 3 unknowns that would change the decision, each with how to find out (win-loss interviews, a product teardown, customer calls, public filings), an owner, and a date. Then the decision: engage, avoid, or learn first, with a review date. End with next actions, not a summary of Sun Tzu.
## Output format
```markdown
# Know the enemy and yourself: [us] vs [competitor]
**Contest:** [customers or segment] · **Decision:** [what this serves] · **Checked:** [date]
## Five factors
| Factor | Us | Them | Edge | Sources |
|---|---|---|---|---|
| Purpose (道) | [fact or unknown] | [fact or unknown] | [Us / Them / Even / Unknown] | [tags] |
| Timing and conditions (天) | | | | |
| Terrain (地) | | | | |
| Leadership (將) | | | | |
| Method (法) | | | | |
## Seven comparisons
| Question for this contest | Edge | Source |
|---|---|---|
| [whose direction do customers and staff back?] | [Us / Them / Even / Unknown] | [tag] |
## How much we know
Unknown edges: [n] of 5 factors, [n] of 7 comparisons
Unknown on our side: [list] · on theirs: [list]
Case: [we know both / we know only ourselves / we know neither] — [one sentence why]
## Their solid and empty points
| Their point | Solid or empty | Evidence | Our matching strength |
|---|---|---|---|
## Where to engage, avoid, or learn first
- **Engage:** [where] — because [their empty point + our strength + terrain]
- **Avoid:** [where] — because [their solid point or our unknowns]
- **Learn first:** [where] — depends on [unknown]
Conditions for victory not met: [none, or which and why]
## Learn first
| Unknown | How to find out | Owner | By |
|---|---|---|---|
## Decision
[Engage / avoid / learn first] — [one sentence]. Review on [date].
```
## Example
**Input:** "We're Pantry Ledger, inventory software for independent cafés. Competitor: StockStack. We sell to cafés with one to three sites in the Kansai region. Decision: should we build supplier ordering this quarter to compete head-on? Facts: our price is $29 a month, theirs is $49 a month [their pricing page]. They have supplier ordering; we don't [their site]. G2 reviews of StockStack mention slow support replies, 'two days or more' in 6 of their last 20 reviews. Our win-loss: won 7 of 9 deals against them in single-site cafés, lost 4 of 5 at cafés with three sites, all citing supplier ordering. Our team: 6 people, 2 engineers. Runway 18 months. Their team size: LinkedIn shows about 40 staff. Their funding: unknown. We don't know our own churn reasons; we never asked. Japan's invoice rules changed last year and cafés asked us about receipts."
**Output:**
```markdown
# Know the enemy and yourself: Pantry Ledger vs StockStack
**Contest:** independent cafés with 1–3 sites in Kansai · **Decision:** build supplier ordering this quarter? · **Checked:** [date]
## Five factors
| Factor | Us | Them | Edge | Sources |
|---|---|---|---|---|
| Purpose (道) | Win 7 of 9 single-site deals; churn reasons unknown | Reviews complain about support speed | Unknown | [win-loss], [review]; [check: our churn reasons] |
| Timing and conditions (天) | Cafés ask us about invoice-rule receipts | Unknown | Unknown | [internal]; [check: StockStack's invoice support] |
| Terrain (地) | Strong in single-site cafés | Strong in 3-site cafés | Split by café size | [win-loss] |
| Leadership (將) | Unknown | Unknown | Unknown | [check: both sides' record of decisions] |
| Method (法) | 6 people, 2 engineers, 18 months runway | About 40 staff; funding unknown | Them on capacity | [internal], [LinkedIn, unverified] |
## Seven comparisons
| Question for this contest | Edge | Source |
|---|---|---|
| Whose customers would recommend them? | Unknown | [check: ask 10 customers] |
| Whose team answers customers faster? | Us, likely | [review]; [check: our median reply time] |
| Who holds the better ground? | Us in single-site, them in 3-site | [win-loss] |
| Whose processes run as stated? | Unknown | [check] |
| Whose resources are larger? | Them | [LinkedIn, unverified] |
| Whose team is more skilled at this product? | Unknown | [check] |
| Whose incentives point at the customer? | Unknown | [check] |
## How much we know
Unknown edges: 3 of 5 factors, 4 of 7 comparisons
Unknown on our side: churn reasons, support reply time, leadership record · on theirs: invoice-rule support, funding, leadership record
Case: we know neither (closer than we assumed) — we can't say why our own customers leave, and half of their side is guesswork.
## Their solid and empty points
| Their point | Solid or empty | Evidence | Our matching strength |
|---|---|---|---|
| Supplier ordering | Solid | [site], 4 of 5 lost deals | None yet |
| Support speed | Empty | 6 of 20 recent reviews | Small team close to customers [check: our reply time] |
| Price for single-site cafés | Empty | $49 vs our $29 | Lower price |
| Invoice-rule receipts | Unknown | — | Customers already ask us |
## Where to engage, avoid, or learn first
- **Engage:** single-site cafés — their price and support gaps meet our price and closeness, on ground where we win 7 of 9.
- **Avoid:** three-site cafés this quarter — supplier ordering is their solid point, and 2 engineers can't match it fast.
- **Learn first:** invoice-rule receipts — could be an opening, but we don't know whether StockStack already handles it.
Conditions for victory not met: "knowing how to use a small force" — building supplier ordering would spend most of our engineering on their strongest ground.
## Learn first
| Unknown | How to find out | Owner | By |
|---|---|---|---|
| Why our customers churn | Call the last 8 churned cafés | [owner] | [date] |
| Our median support reply time | Pull from the help desk | [owner] | [date] |
| StockStack's invoice-rule support | Trial account and their docs | [owner] | [date] |
## Decision
Avoid head-on supplier ordering this quarter; engage in single-site cafés on price and support, and learn the three unknowns first. Review on [date].
```
## Checks before you send
- [ ] The contest and the decision are named in one sentence each.
- [ ] Every fact on both sides has a source tag; nothing about either side is invented.
- [ ] Unknowns are written "unknown" with a "[check: ...]" saying how to find out, on our side as well as theirs.
- [ ] The output says which of the three cases (both, only ourselves, neither) applies, and why.
- [ ] Each "engage" pairs a rival's empty point with a strength of ours that has evidence.
- [ ] It ends with a decision, a learning plan with owners and dates, and a review date.
- [ ] Any line of the Art of War is our own rendering, not a published translation.
## Common mistakes
- Studying only the rival. The line's middle case, knowing yourself but not them, is a coin flip; knowing them but not yourself is no better. Self-knowledge is half the table.
- Reading the line as "you will win every battle". 殆 means danger: the promise is that you won't be caught in a fight you misjudged, which includes choosing not to fight.
- Filling unknown cells with impressions. "They're probably slow" is not a fact; mark it and find out.
- Attacking the rival's strongest point because it's the most visible feature. Water goes around the high ground.
- Treating the five factors as decoration. If a factor doesn't change the decision, say so in one line and move on.
Share it with your team
Add it to an AI Skill Pod and everyone on the team, contractors included, uses the same version from their own AI. Read the best ways to share AI skills with a team
Changelog
- v1.0.1 · · Description: the five factors now carry the names the steps use (purpose, timing and conditions, terrain, leadership, method), not "discipline" for method.
- v1.0.0 · · First release.
Sources
- 孫子兵法 (original Chinese text) - Wikisource (read 2026-11-04)
- Ritsumeikan University, College of Letters: faculty column on the Sunzi and the Yinqueshan bamboo slips (read 2026-11-04)
- The Art of War - Literary Themes for Students (Gale), Encyclopedia.com (read 2026-11-04)