phone-prep
Skill1 starsChanged 4 months agoMITView on GitHub
Build a one-page phone brief — goal, script, questions, comebacks — before the user calls a billing office, insurer, or collector. Also handles the post-call debrief.
In this repository11 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- benchmark-priceSkill
- collectionsSkill
- draft-emailSkill
- financial-assistanceSkill
- insurance-appealSkill
- intakeSkill
- review-billSkill
- send-letterSkill
- strategySkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
Write cases/<case>/call-brief-yyyy-mm-dd.md, phone-screen friendly (the user will read it mid-call):
- Distinctive: line 20
Always first: "Who am I speaking with, and what's a reference number for this call?"
- Distinctive: line 27
Always last: "Please note the account is in dispute.
- Very specific: line 43
Ask what happened, fill the capture sheet into call-log.md, then immediately draft the same-day written recap email (draft-email): "Per our …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Phone prep
You can't make the call; you make the caller unbeatable. Billing reps do this all day — the user does it once, scared. The brief closes that gap.
Before the call: the one-page brief
Write cases/<case>/call-brief-yyyy-mm-dd.md, phone-screen friendly
(the user will read it mid-call):
- Goal — the single thing this call must produce (usually: a name, a fact, or a written follow-up commitment — almost never a payment).
- Opening statement — 2–3 sentences, verbatim, with account number.
- The questions — 5–10, ordered, with blanks to fill:
- Always first: "Who am I speaking with, and what's a reference number for this call?"
- Then the case-specific ones (examples): "What is the status of my
itemized-statement request from
<date>?" / "Is this account on hold while my financial-assistance application is pending?" / "What discount is available if this is settled as a lump sum?" / "Can you email me that in writing today?" - Always last: "Please note the account is in dispute. What did you record?"
- Likely pushback + comebacks — pull the 3–4 relevant entries from
knowledge/objections.md, quoted ready-to-say. - Do-not lines — what the user must NOT do on this call. Default: don't agree to any payment or plan, don't state income figures to a collector, don't say "I owe"; say "the amount claimed." For collectors: don't confirm the debt is theirs before validation.
- Capture sheet — name / department / reference # / promises / deadlines.
Rehearse the opening with the user once if they want. Remind them: polite, brief, endlessly repetitive is the winning style — reps escalate persistent callers and stonewall angry ones.
After the call: debrief
Ask what happened, fill the capture sheet into call-log.md, then
immediately draft the same-day written recap email (draft-email):
"Per our call today with <name>, reference #<n>, you confirmed…" — the
recap converts a phone promise into evidence. Update deadlines in case.md
and name the next action.
Pre-care calls (no bill yet)
For upcoming appointments/procedures, the brief is the four prevention
questions in knowledge/before-care.md (network status for the specific
visit type, financial-assistance application, where tests/referrals get
processed, best-price-if-paying-today) plus a GFE request when self-pay.
Same capture rules: names, dates, and answers into a dated note the user
keeps.
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.

