RUNS + STATS · LAST 30 DAYS
Dashboard
The triage view. Open signals at the top, pending operator approvals beside them, then runs in reverse-chronological order with verdicts and phase completion at-a-glance.
28
Runs · last 30 days
82%
Pass rate
7
Open signals · high
9 normal in queue
—
Pending approvals
queue empty
SIGNAL · HIGH
SIG:lob-webhook-broken — Lob webhook target points at deleted Vercel projec…
SIGNAL · HIGH
SIG:journal-notebook-leakage — Journal/notebook/currency artifacts conceptually l…
5× CAPSTONE
conductor → architect · journal split · RUN:007
PATTERN · 3 RUNS
PAT:last-mile-handoff — Last-mile handoff
NEW RUN
RUN:sample-001 · Phoenix Lunch And Learn Targets
Recent runs
#031
Master Reset Rebuild Audit
PASSThree structural problems matter more than any individual item: 1. Drift between MASTER_PLAN and live state. At least 5 FIN items in the "still open" bucket are already done. FIN-1
62 items × 11 agents = ~660 sub-outputs with massive duplication; scoped to 5 specialist angles + audit-quality. Conductor-synthesized. Precedent: run #028.
#029
Ewing Book Report
PASS34 net-new systems in the 7-day window across 235 commits 3 swarm-build runs closed (#018 Gmail + cost ledger, #019 DPP/Wieser consolidation + provenance, #020 audio corpus + trans
Full roster — Ewing explicit "have agents vote" required all 11 votes.
#028
Vercel Page Audit
PASSThe May 4 work was issued a PASS but only 6 of its 12 named deliverables actually existed on disk. The four missing ones were all HTML rollups of analysis that did get done — the a
Internal infrastructure audit. Excluded hunter (no people-finding), market-analyst (no numbers/market math), writer/draper/tech-translator (no client-facing copy). Conductor-synthesized Phase 2/3 due to strong convergence across the 4 working agents — narrow divergences (Supabase scope, kill-vs-converge axis, HOLD definition) resolved without a second-pass cross-read.
#027
Email Template Audit
PASSCustom Ewing-specified roster. market-analyst role conductor-synthesized; architect role covered by Phase 0 DB probe; quarterback/draper/storyteller skipped — assignment is data→template gap analysis, not strategic positioning.
#022
Deal Page Consolidation
PASS420+ files across 7 different storage layers. The richest operational data (call intelligence, target company research, meeting decisions) is trapped in static HTML files that no o
Full roster (all 11)
#021
Ui Ux Purpose Audit
FIX-AND-SHIPThe system covers 62% of table-stakes CRM features with strong operational depth (deal pipeline, meeting intelligence, caller, letters, deal rooms). But it grew from 0 to 75 pages
internal-only — no external people/company discovery; pure UI/strategy audit
#020
Exa Consolidation
PASSWe use about 40% of what Exa offers. We call one endpoint out of six available. On that one endpoint, we pass 8 of 22 parameters. The biggest gap is Websets — Exa's structured data
The build was intern-scoped wiring (API routes, component, feature defs, recipes). The audit was a single-agent deep read against live Exa documentation. Neither required 5+ agents.
#019
Skills Sync Architecture
FIX-AND-SHIPThe architecture is GitHub-as-truth with a hybrid push channel: GitHub Actions writes a manifest row to Supabase on every merge; a LaunchAgent on each machine polls that row every
`internal-only` preset excludes `hunter` — no people-finding required for an infrastructure problem. Phase 2 cross-read agents were not dispatched; Conductor synthesized directly from all 9 Phase 1 journals to avoid further API timeouts.
#017
Buyside Hunt Routine Review
PASSEwing built a routine that turns a buyer's criteria (geography, EBITDA band, verticals) into a polished acquisition target shortlist in 8 minutes. The routine ships an HTML letter
unknown
#016
Letter Sending Product Spec
PASS1. There's already substantial code — 12+ files across Python, TypeScript, and JavaScript — but it's fragmented across 3 overlapping implementations and frozen. The plumbing exists
Full roster
#013
Os Tab Audit
CONCERNSThe OS is a deal machine dressed as a developer dashboard. The two most important revenue tools — the caller app (where SDRs make calls) and deal management (where deals get tracke
Full roster — all 11
#012
Meetings Fireflies Backfill
PASSPopulated 5 `meeting_notes` rows from Fireflies transcripts (last 45 days, NC-participant filter). Each row renders at `os.chapter.guide/meetings/<id>` with all four field groups a
Mechanical backfill task — schema migration + transcript extraction, no strategic synthesis required
#010
Vercel Deep Dive
PASSAudit + rebuild of chapter.guide Vercel deployment. 117-target HR.com consolidation, deal-page template with transcript player, swarm-room polish, v2.0 doc-standards proposal. The 5x bonus event landed here (conductor → architect, v2.0 proposal).
Phase 1 research delegated to 5 specialized subagents; conductor played remaining agent roles synchronously due to session-bound execution model
#009
Agent Swarm Mode Feature System
PASSAudit-only run. 107+ files classified as the Agent Swarm Mode feature system. Notebook quality scored 4/5; consistency 3.5/5. Swarm-upgrade flagged as not safe for unsupervised daily automation — needs 6 specific guardrails.
5-agent subset: pure internal audit + classification — no buyer research, no client-facing copy, no deal context needed
#008
Salesfinity Weekly Pipeline
PASSNull-data run. Zero call records across Salesfinity, Fireflies, SpokePhone, Gmail. Correctly reframed as 'pre-launch state' diagnostic rather than fabricated analysis. Surfaced two persistent stack-activation gaps.
Pipeline ops assignment — no contact discovery (hunter), no emotional copy (draper), no market sizing (market-analyst), no narrative coherence needed (storyteller), no infra implications (architect)
#007
Repo Consolidation One Repo
PASS5× CAPSTONEValidated and refined the 4-repo → 1-repo consolidation plan. ~$12k/year savings, 6-week payback. Three high-risk steps (Lob webhook, deals migration, dirty git tree) calendar-gated to mid-week daytime. Stage 1 inspection discovered the deals migration was already 80% done.
pure-strategy preset — internal infrastructure decision; no contact discovery (hunter excluded), no client emotional layer (draper excluded). Listener covered intern unstated needs in lieu of draper.
#006
Hrcom Data Room Blueprint
PASSProduced complete 38-document data room organized into 8 sections, with summary statements, 'Why This Matters' explanations, numbered fill-in prompts. Built for HR.com (Debbie McGrath) with Paychex flagged as interested buyer.
Data room construction — storyteller, tech-translator, debrief excluded (writer handled plain English + narrative; debrief Phase 6 only)
#005
Migration Audit Confirmed
PASSListener-driven analysis of migration intent across recent transcripts. Stub run — full notebook content not preserved in archive.
Listener-led intent extraction; minimal subset
#004
Enrichment Audit Post Run003
PENDINGOne signal was fully deployed. Three were not. The one that was fixed (caching confirmed sponsors in Supabase) was fixed today — in this session — only because Ewing directly asked
Targeted audit panel — storyteller/quarterback co-leads per Ewing directive; voted in architect (system state), audit-quality (independent verdict), debrief (learning extraction). hunter/writer/draper/market-analyst/listener/tech-translator excluded — no contact discovery, no copy, no market sizing needed.
#003
Deal Tracker Deal Room
PASSShipped deal-room route at /room/[slug] with v_deal_room_content view and deal_plays table. Identified 'three disconnected systems' as core coherence failure. Caught Clay pricing correction ($0.02 → $0.30-$1.00) preventing budget surprise. Designed knowledge_atoms universal table (later flagged as not deployed in RUN:009).
Full strategic + research + code work warrants full slate
#002
Hrcom Buyers Exa Mastery
PENDINGEstablished single canonical ExaClient + 22-column process_runs schema. Surfaced 15 unfollowed Debbie-named warm contacts from in_scope[].debbie_note_raw. Identified L&D / credentialing buyer cohort (Pearson, LRN, Skillsoft, Vector, Relias). Conductor recovery script reconstructed 12 dossiers from process_runs after Stage D crash.
Broad strategic + research + code work warrants full slate
Open signals · high
SIG:lob-webhook-brokenhigh
Lob webhook target points at deleted Vercel project
2-run evidence · stage 2/4
SIG:journal-notebook-leakagehigh
Journal/notebook/currency artifacts conceptually leak across 9 runs
1-run evidence · stage 2/4
SIG:runs-006-008-missing-archiveshigh
Runs 006-008 missing from public/swarm/ + dashboard RUNS array
1-run evidence · stage 2/4
SIG:swarm-upgrade-threshold-bypasshigh
Swarm-upgrade applied 15 single-run proposals despite 3+ documented threshold
1-run evidence · stage 2/4
Service health
SVC:lob
SVC:salesfinity
SVC:fireflies
SVC:supabase
SVC:doppler
SVC:exa
SVC:vercel
Pending approvals
No pending proposals · queue empty