message-writer
Skill2 starsChanged 56 days agoMITView on GitHub
Ghostwriter — writes the LinkedIn invitation note, the opening message once connected, and a variant B, built on each lead's why-now sentence, then runs human-voice on them. Use for "write the messages", drafts for a batch of leads, or fixing a message that sounds automated.
In this repository20 other files it gives its agents.
- AGENTS.mdAGENTS.md
- buying-committeeSkill
- call-readySkill
- campaign-launchSkill
- command-centerSkill
- competitor-watchSkill
- fit-checkSkill
- human-voiceSkill
- intent-radarSkill
- market-mapSkill
- morning-briefSkill
- nudgeSkill
- outbound-briefSkill
- post-studioSkill
- priority-rankSkill
- reach-checkSkill
- reply-deskSkill
- social-warmupSkill
- what-worksSkill
- why-nowSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 10
Write from the why now sentence and the proof in the lead card, in the voice set by outbound-brief.md.
- Very specific: line 27
When why-now is thin, write shorter rather than padding it.
- Distinctive: line 31
Read the lead card: why now, proof, seat, quality.
- Distinctive: line 32
Write the three drafts.
- Very specific: line 33
Run human-voice on all of them.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Message writer — Ghostwriter
House rules: ../command-center/shared/house-rules.md
Write from the why now sentence and the proof in the lead card, in the voice set by outbound-brief.md. Use the lead's language if Pastel shows it; if not, the user's.
Three drafts per lead
| Draft | Length | Job |
| Invitation note | ≤ 180 characters, counted | Get the connection accepted. The why-now point and a light question. No pitch, and no product or company name of ours. |
| Opening message | ≤ 50 words | Sent once connected: the why-now point, what it means for them, one question they can answer in a line. |
| Variant B | same limits | Same facts, different entry point, so the user can pick or A/B test. |
The Buying Committee seat changes the angle: deciders hear outcomes and cost, users hear about their day, champions hear back what they already said.
Rules
- Open with the topic, not the surveillance: talk about what the post they engaged with was about ("teams juggling five outbound tools…"), never "I saw your comment" or "you commented DEMO".
- The ask is small: something they can answer from their phone in one line.
- The Buying Committee seat, if any, decides the angle (see above).
- When
why-nowisthin, write shorter rather than padding it.
Steps
- Read the lead card: why now, proof, seat, quality.
- Write the three drafts.
- Run
human-voiceon all of them. - Return them as drafts;
campaign-launchhandles sending after approval.
Output
# Drafts — [date]
### [Person] · [Company] · priority [n]
Why now: …
Invitation note ([n]/180):
> …
Opening message:
> …
Variant B:
> …Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.