path-to-close

name	path-to-close
description	Generate the founder's path-to-close brief — round objective, current state, pipeline strategy, relationship paths, and next best moves, worked backward from the funded round. Use when the founder asks "how do I get this round done," "where am I," "what's my path," or weekly to refresh the command center.

Path to Close

Work backward from a funded round and tell the founder what moves matter now — strategy, not a status report.

Orientation (always first)

1. If the belief-engine skills are available, run /beliefs raise to load the current belief state, then /open-decisions. If they aren't installed, skip them — the progress/ and round status reads in steps 2-3 are the durable substitute.
2. Read: round status, round investors list, round tasks list, round dataroom status.
3. Read the narrative and terms docs in company/ (e.g. company/profile.md), the 2-3 most recent progress/ entries, and the investor MDs for every tier-A target plus anyone with status Met or later in the funnel (Met, Diligence, Term sheet, or Committed).

Delegate

• Spawn the process-auditor subagent — its findings become the honest core of "Current State" and "Gaps & Risks". Never write "Current State" from optimism; write it from the audit.
• For the top 3-5 tier-A targets with no warm path, spawn investor-researcher to map candidate intro paths (one agent per investor, in parallel).

Compute the gap math (do not skip)

• Open allocation = target − committed. State it in dollars.
• Weight the live pipeline by stage (rough, illustrative conversion: Committed 90%, Term sheet 75%, Diligence 50%, Met 25%, Meeting set 15%, Replied/Outreach sent 5%, Warm path building 2%, Not started 0%) → expected value of current pipeline.
• Residual gap = open allocation − pipeline EV. New-meeting demand comes from this residual, not the full open allocation — otherwise you double-count the pipeline you already have.
• Residual gap → "you need roughly N new lead-quality first meetings" = residual ÷ (avg check × meeting→commit rate), assuming ~20% meeting→commit for tier-A fits. Show the arithmetic so the founder can argue with it.
• Tranche/SAFE-cap awareness: note allocation remaining per cap tier if terms define tranches.
• Staleness: flag every tier-A investor with lastContactAt older than 14 days or empty.

Output

Write progress/YYYY-MM-DD-path-to-close.md (today's date) with exactly these sections:

1. Round Objective — what future we are making true (amount, close date, composition: lead + angels).
2. Current State — what is true right now. Sourced from the process audit, not vibes. Include the gap math.
3. Pipeline Strategy — what needs to become true: how many lead-quality targets, warm intros, high-conviction meetings, and angels, vs. what exists.
4. Relationship Paths — for each priority target: the path (who → who → partner), warmth, and what would strengthen it.
5. Next Best Moves — causal moves, not generic tasks. Each move says what it changes ("converts believer X into intro node", "kills objection Y with proof Z"). Include explicit STOP-doing items.
6. Gaps & Risks — what could quietly kill the round: coverage gaps, proof-asset debt, stale relationships, momentum risks.

Close the loop

• Propose 3-5 round tasks add for this week's moves (one per move, with due dates). Propose only — the founder approves in the app.
• If /observe is available, run it on the finished brief so the strategy enters the belief state; otherwise skip.
• End with a short chat summary: the gap number, the single highest-leverage move, and what changed since the last brief (if one exists in progress/).

Rules

• All structured writes go through round CLI proposals; never --yes unless told.
• File writes are captured for founder review — that's success; don't retry.
• If the data contradicts the founder's stated plan, say so in Gaps & Risks. This brief is worthless if it flatters.