agentleFS
TrustPagerTrustPager/Business_Operating_System/skills/missed-call-recovery/SKILL.md

missed-call-recovery

Skill8 starsChanged 5 months agoMITView on GitHub

Find recent missed calls, identify the caller, and draft a recovery SMS or callback message — one per missed call, ready to send.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the fetch script. It returns every missed inbound call from the last 24h (configurable), enriched with:

    line 30
  2. Distinctive:

    Use the contact's preferred channel if it's set on the record.

    line 58
  3. Distinctive:

    Wait for explicit yes/no per message.

    line 64
  4. Distinctive:

    No fabrications: If a missed call doesn't have a name attached, the draft must say "the number that called" — never invent a name.

    line 71
  5. Distinctive:

    Quiet hours: If the current time is before 7am or after 8pm in the recipient's timezone (or unknown), draft an email instead of SMS.

    line 73
  6. Distinctive:

    One recovery per number: If the same number called multiple times and we've already recovered the first, treat the rest as resolved.

    line 74

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/missed-call-recovery

When someone misses a call, the longer the gap before a response, the lower the chance of recovery. This skill makes that gap as short as possible — pulls every recent missed call, looks up who it was (existing contact? opportunity? cold caller?), and drafts a per-call recovery message you can send with one approval. For locally bought businesses, answer speed is the first rung of the whole acquisition ladder, and usually IS the constraint (business-method.md §10.5): first responder wins the job.

Step 1 — Pull the data

Run the fetch script. It returns every missed inbound call from the last 24h (configurable), enriched with:

  • Phone number that called
  • Whether we have a contact for that number
  • Linked opportunity (if any) + current stage
  • Whether the caller has been called back already since the missed call
  • Time since the missed call
python ~/.claude/bos-run.py missed-call-recovery

Pass --hours 48 for a longer window, or --include-callbacks to also surface calls that were already recovered (useful for reviewing the day's recovery work).

Fallback if the script can't run (auth/network): say so briefly, then pull the window by hand — mcp__trustpager__list_phone_call_logs, plus get_contact_deals per matched caller — and triage from that. Proceed with what you have.

If TrustPager isn't connected at all: say so plainly, then offer the keyless path — the owner reads you the missed numbers (or pastes their phone's call list) and you draft the recovery texts right in chat for them to send from their own phone.

Step 2 — Triage and draft

For each missed call in the output, present a one-line summary to the user, then propose the recovery action:

Caller typeDefault recovery
Existing contact with open opportunity"Sorry I missed your call — was about [opportunity name]?" SMS, then offer to schedule a callback.
Existing contact, no open opportunityFriendly callback SMS. Ask what they were calling about.
Unknown number (no contact)SMS asking if they were trying to reach the business. Do NOT create a contact yet — wait for a reply.
Caller already recoveredSkip silently. Don't ask the user.

Use the contact's preferred channel if it's set on the record. Otherwise default to SMS for missed calls (faster than email).

Step 3 — Send with approval

For each drafted message:

  • Show the user: who, the phone number, the proposed message, the channel.
  • Wait for explicit yes/no per message. NEVER batch-send.
  • On yes: send via mcp__trustpager__send_sms (or send_email if email is the channel).
  • On no: ask what to change (tone? length? skip?), or move on.

Important behaviours

  • Content guardrails. Customer-facing copy uses no em dashes, invents no facts, quotes, or numbers, and names no third-party vendor. Write it in the owner's brand voice; the framing and marketing psychology are the owner's choice. The rules are in knowledge/content-rules.md. The register for these messages is knowledge/communication-voice.md: plain, warm, short.
  • No fabrications. If a missed call doesn't have a name attached, the draft must say "the number that called" — never invent a name.
  • No batching. Each send is its own approval.
  • Quiet hours. If the current time is before 7am or after 8pm in the recipient's timezone (or unknown), draft an email instead of SMS.
  • One recovery per number. If the same number called multiple times and we've already recovered the first, treat the rest as resolved.
  • Don't open with "Sorry I missed your call" if they called more than 6 hours ago. That feels insincere. Lead with "Hey, I saw your call earlier — what's up?" instead.
  • Make it standing, not manual. If the operator runs this more than occasionally, the durable fix is an automatic missed-call text-back; offer /automate-this to wire it (§10.5, with the 60-second first touch as the aspiration, §10.3, directional). On the keyless path (TrustPager not connected), skip the automation offer.

Output shape

The skill should end with a one-line summary: "Recovered N of M missed calls. K already had a callback. R skipped."

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.