## TL;DR
The line is contribution first, self-interest second, judged by pattern not by post. If most of your activity helps people with no benefit to you, occasional relevant mentions are contribution; if most of your activity points at your tool, every post is spam regardless of quality. Apply the neutral-expert test to each mention and keep the ratio heavily toward pure help.

```text
where the line is between contribution and spam
```

## Use this when
- You are briefing an agent on how much self-promotion is acceptable
- A post got flagged and you need to judge whether it was fair
- You are reviewing your fleet's participation for drift toward spam
- A community member questioned your motives
- You are writing the self-promotion section of a participation policy

## Not for this skill when
- You are appealing a specific ban (see the ban appeal skill)
- You want tactics to promote without getting caught (this skill is the opposite of that)
- The venue has explicit written rules (those override any general line)
- The question is about email or ad spam, not community participation

## Steps
1. Apply the neutral-expert test to every mention before posting. Would a neutral expert with no stake in your tool recommend it in this thread? Yes means the mention serves the asker; maybe means it serves you; no means it is spam. Only yes ships.
   ```text
   thread: [link or id]
   neutral expert would recommend: [yes / maybe / no]
   decision: [post | skip]
   ```
   Expected output: a test result per planned mention. Success check: you skip more mentions than you post.

2. Keep the ratio heavily toward pure contribution. As a rule of thumb, at least four link-free helpful replies for every one that mentions your tool, sustained over time. Ratios are judged by pattern; one promotional burst resets the goodwill of a month.
   ```text
   last 30 days: [n] helpful replies, [m] mentions, ratio [n:m]
   ```
   Expected output: a tracked ratio. Success check: the ratio stays above 4:1 and mentions are spread out, not clustered.

3. Check the relevance bar per mention, not per account. A trusted account posting an irrelevant mention still spams that thread; a new account posting a perfect answer still contributes. Each post is judged where it lands.
   ```text
   [ ] the mention answers the specific question asked
   [ ] no better non-self answer was skipped to make room for it
   [ ] the mention is brief; the help is the post, not the pitch
   ```
   Expected output: three checks per mention. Success check: removing your tool's name still leaves a helpful reply.

4. Watch for the pattern signals moderators watch for. New account plus first post with a link, links in every reply, replying only to threads where your tool fits, and never engaging outside your niche all read as spam regardless of content quality. Audit your own pattern monthly.
   ```text
   [ ] replies span topics beyond my tool's niche
   [ ] some threads engaged with no possible self-benefit
   [ ] no link in the first several replies of any new account
   ```
   Expected output: a pattern you would defend publicly. Success check: a moderator reviewing your history sees a member, not a campaign.

5. When in doubt, ask the moderators before posting. A quick "is this kind of mention welcome here" message to the mod team takes a day and buys permanent clarity. Communities respect askers far more than they respect assumers.
   ```text
   asked mods: [date], answer: [summary]
   ```
   Expected output: written guidance from the venue. Success check: you can quote the mod team's answer when anyone questions a mention.

## Variant phrasings
### How much self-promotion is allowed in forums
Allowance phrasing. Steps 1 and 2: the neutral-expert test per mention and the 4:1 ratio over time.

### Why was my forum post flagged as spam
Diagnostic phrasing. Walk steps 3 and 4 against the flagged post: relevance where it landed, and the account's pattern around it.

### Self-promotion rules for devtool marketing
Marketing phrasing. Same line, framed for marketers: contribution is the strategy, mentions are the exception.

### Contribution vs promotion in developer communities
Conceptual phrasing. The whole skill: judged by pattern, the mention must survive the neutral-expert test, ratio does the rest.

## Why it happens
Moderators cannot read intent, so they read patterns: ratios, timing, relevance, and breadth. A single post is ambiguous; fifty posts are a portrait. The line feels blurry post-by-post because it is drawn across the pattern, which is also why "but this one post was helpful" never wins an appeal. Contribution-first works because it builds the pattern that makes occasional mentions legible as help.

## Edge cases / pitfalls
- Disclosure does not move the line. Saying "I am an agent" makes spam honest spam; the pattern still has to be contribution-first.
- Highly relevant mentions can still cluster badly. Five perfect mentions in one day reads as a campaign; spread them out.
- Competitor threads are a trap. Mentioning your tool in a thread about a competitor fails the neutral-expert test almost always; answer helpfully without the mention or skip the thread.
- The ratio is a floor, not a target. 4:1 keeps you safe; 10:1 makes you welcome. Aim past the line, not at it.

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_WTv43VgsZgo0yrRQIMtHow