multithread
Skill2 starsChanged 24 days agoMITView on GitHub
Map the real buying committee on an account or deal and get multithreaded across it. Use this skill whenever a seller says "multithread this account", "map the buying committee", "who else should I be talking to at [account]", "this deal is single-threaded", "widen this deal", "get more people involved", "map the stakeholders", "build the org chart for this deal", or is riding a deal on one contact that needs more coverage. It works from two starting points: a brand-new account with no coverage, or an existing live deal running on a single thread that needs to be widened before it stalls. It pulls context from Loop first (falling back to CRM, call transcripts, and email), maps the committee by role, writes a value hypothesis per stakeholder, and drafts the outreach to reach each one. Prefer this over guessing who to email next any time a deal depends on more than one person.
In this repository10 other files it gives its agents.
- account-planningSkill
- call-follow-upSkill
- call-prepSkill
- champion-enablementSkill
- competitive-displacementSkill
- deal-advancementSkill
- manage-upSkill
- one-page-business-caseSkill
- qbr-prepSkill
- sales-playsSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 4 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
New account: no coverage yet. You are choosing 3 to 5 entry points cold.
- Very specific: line 39
Loop first: If the Loop MCP (https://mcp.fluint.io/mcp) is connected, use it first.
- Distinctive: line 43
No Loop? Detect what is live and fall back: Do not assume a vendor. In parallel:
- Distinctive: line 45
Call transcripts (Salesloft, Gong, Clari, Chorus, Fathom, Fireflies.
- Very specific: line 58
Read references/frameworks.md (Buying committee + shadow org).
- Very specific: line 66
Read references/frameworks.md (which-means).
- Very specific: line 73
Read references/output-templates.md and use the exec-level multithreading email patterns.
- Distinctive: line 102
Three conversations before any group meeting: No surprises in the room.
≈ “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
Multithread
What this does
Single-threaded deals are the highest-risk deals in the pipeline. One person gets promoted, goes quiet, or gets overruled, and the deal dies. This skill widens the coverage:
- Maps the real buying committee: who decides, who they listen to, who would block this.
- Writes a value hypothesis per stakeholder, tied to the metric that person owns.
- Drafts the outreach to reach each one, easy for a champion to forward.
- Sequences it: who to reach first, and who to reach through the champion versus direct.
Two starting points, same map:
- New account: no coverage yet. You are choosing 3 to 5 entry points cold.
- Existing deal, single-threaded: you have one contact. You are widening through them before the deal stalls on one person.
The deliverable is a multithreading plan the seller can act on today.
Step 1: Lock the target and pull context
Loop first. If the Loop MCP (https://mcp.fluint.io/mcp) is connected, use it first. Pull the
account, the deal if one exists, and the contacts and meeting history. Loop already
fuses CRM, calls, and emails, so it is the fastest way to see who is already in the deal.
No Loop? Detect what is live and fall back. Do not assume a vendor. In parallel:
- CRM (HubSpot, Salesforce, Attio, Pipedrive) = existing contacts, roles, deal history.
- Call transcripts (Salesloft, Gong, Clari, Chorus, Fathom, Fireflies; or via Clay / Deepline) = who has been named or mentioned on calls.
- Email (Superhuman, Gmail, Outlook) = who is cc'd, who forwards to whom.
Set the entry point: new account, or an existing deal riding on one thread. If existing, identify the single contact you have and whether they are a champion (see below).
Missing a source? Flag it, and make the gap a move. Mark it inline as [MISSING: ...] and continue. Every gap gets the single highest-leverage way to close it: the one question, to the one person, at the next touchpoint. Not "get more data."
Who can actually say yes? Name the people with the power to approve this, or to block it, and mark whether they are engaged yet. Power is often not in the room. If the deciders are absent or untouched, that is the headline, not a footnote.
Step 2: Map the buying committee
Read references/frameworks.md (Buying committee + shadow org). Map the real decision
structure, not the org chart. Identify 3 to 5 people across the roles that recur: Champion,
Economic Buyer, Technical Evaluator, Internal Skeptic, and the SLIP stakeholders (Security,
Legal, IT, Procurement) who can stall a deal late. For each, capture their likely role in the
decision and what they care about most.
Step 3: Write a value hypothesis per stakeholder
Read references/frameworks.md (which-means). For each person, write one sentence that would
land with their specific priorities, tied to what is happening at the account right now. Not
generic. Chain it to their metric: this user's pain, which means this manager's number, which
means this exec's KPI.
Step 4: Draft the outreach
Read references/output-templates.md and use the exec-level multithreading email patterns.
Draft a short message per stakeholder, under 100 words, leading with relevance. For a new
account, these are entry points. For an existing deal, use invitational language to widen
through your champion ("is this a focus for them? I will loop them in for feedback"). Keep the
tone easy to forward, so a buyer-side exec can pass it across their team.
Step 5: Prioritize and sequence
Recommend who to reach first and why, then how to sequence the rest by likelihood to engage and strategic value. Note who you approach directly versus through the champion. Two rules from the frameworks: confirm at least three independent conversations before presenting to a group, and never run pricing in a group.
The deliverable
Produce the multithreading plan using the skeleton in references/output-templates.md:
- Buying-committee map (name, role, real influence, disposition, contact status)
- Value hypothesis per stakeholder (the one-line soundbite tied to their metric)
- Drafted outreach (one short message per stakeholder, ready to send)
- Sequence (who first and why, direct vs through the champion)
The operator's read
Before you hand it over, add the part a summary cannot: the one thing an experienced operator would notice that the data does not show on its own. The non-obvious risk. The stakeholder who matters more than their title. The deal that is not what it looks like. The strength that is quietly a trap. One or two sentences of judgment, called out plainly. Anyone can produce the map. This is the read.
Operating principles
- Single-threaded is the highest-risk deal you have. Widen before it stalls.
- Map communication flow, not the org chart. Influence, not title.
- Three conversations before any group meeting. No surprises in the room.
- Never run pricing in a group. Some conversations are one-on-one only.
- Voice guard. Anything the buyer will read opens with their words, from the call or their own materials, in the first line. Banned openers: "Great connecting," "As discussed," "I wanted to follow up," "Hope this finds you well," "Circling back." If a sentence would survive being pasted into a different deal, it is too generic. Cut it.
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.

