batch-draft
Skill1 starsChanged 2 months agoMITView on GitHub
Draft cold emails to several companies in one session for one person, each individually researched and never templated. Use when a batch of outreach is planned. Respects the five-per-day cap and stops early if research quality drops.
In this repository37 other files it gives its agents.
- wsb-data-pipelineSkill
- docx-exportSkill
- harvard-mdxSkill
- humanise-docx-reportSkill
- make-pdfSkill
- bom-builderSkill
- ring-testSkill
- cold-emailSkill
- cv-intakeSkill
- cv-upgradeSkill
- find-contactSkill
- followupSkill
- humanize-emailSkill
- log-outreachSkill
- outreach-statusSkill
- target-researchSkill
- erc-statusSkill
- log-sessionSkill
- score-auditSkill
- submission-checkSkill
- sumo-statusSkill
- trial-runSkill
- vr-statusSkill
- academic-registerSkill
- aviation-data-sourcesSkill
- business-frameworksSkill
- consultancy-registerSkill
- excel-deliverablesSkill
- forecasting-methodsSkill
- operations-frameworksSkill
- secondary-data-sourcesSkill
- archive-indexSkill
- event-proposalSkill
- meeting-minutesSkill
- read-boxnoteSkill
- captureSkill
- new-scriptSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 24
For each company, run the full
/cold-emailpipeline: No shortcuts. In particular, no reusing paragraph 1 with the company name swapped — … - Distinctive: line 26
Cross-check the batch before finishing: Read all the drafts together:
- Distinctive: line 63
Every draft gets its own research.
- Distinctive: line 64
If two drafts share a sentence, both get rewritten.
- Distinctive: line 65
Never fill a count with a weak draft.
- Distinctive: line 66
Never batch across both people in one command.
≈ “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
/batch-draft, several companies in one pass
Usage: /batch-draft "<Person Name>" --tier 1 --count 5
Batch means several emails written in one session. It does not mean one email sent to several people. Every draft gets its own dossier, its own contact, its own match and its own angle, or it is not worth sending.
Preconditions
- The person's profile is complete, no
TBD shared/targets.mdhas enough untouched companies at the requested tier--countis 5 or fewer. The cap exists because research quality collapses past it, and research quality is the only reason cold email works
Steps
- Pick the companies. Tier 1 first. Skip anything already in the tracker, anything the other person contacted within the last week, and anything in
no-replyfrom within three weeks. - Research in parallel, draft in series. Spawn one
company_researcherper company at once — the searches are independent and this is where the time goes. Then work through the drafts one at a time, because each needs its own attention and parallel drafting is how templates creep in. - For each company, run the full
/cold-emailpipeline. No shortcuts. In particular, no reusing paragraph 1 with the company name swapped — that is the exact failure this skill is most likely to produce. - Cross-check the batch before finishing. Read all the drafts together:
- Does any sentence appear in two of them? Rewrite it. Identical phrasing across drafts means the batch was templated
- Does the same project lead in all of them? Sometimes correct, sometimes a sign the match step was skipped
- Are the subject lines meaningfully different from each other?
- Stop early if quality drops. Three well-researched drafts beat five thin ones. If a company's dossier came back with nothing specific, drop it from the batch and say why rather than writing a generic email to fill the count.
- Update the tracker with one row per company.
- Recommend a send order and schedule. Best-fit company first, while attention is freshest on the reading side. Tuesday to Thursday morning. Spread across days rather than five in one hour, so replies can be handled properly as they arrive.
Output format
## Batch: <person>, <n> drafts
| # | Company | Contact | Verified | Angle | Words | Status |
|---|---|---|---|---|---|---|
**Cross-check**
- Repeated phrasing: none | <what was rewritten>
- Lead project variety: <which project led where>
- Subject lines distinct: yes | no
**Dropped from the batch**
- <company> — <why>
**Send order**
1. <company> — <day> — <why first>
**Blocked**
- <company> — <what is missing>Rules
- Five drafts maximum per person per session.
- Every draft gets its own research. A dossier reused across companies is a template.
- If two drafts share a sentence, both get rewritten.
- Never fill a count with a weak draft. Dropping a company is the correct move.
- Never batch across both people in one command. One person per invocation, so their evidence cannot bleed.
- Nothing is sent. All of it lands in
drafts/for a human.
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.

