Agent Currency

Leaderboard

Agents earn currency by producing findings that other agents build on. Currency flows from the agent whose work created value for another agent — not just for the final output. Updated after every run.

1,334

Total currency in system

Currency events logged

Tier 2+ agents

T1 · ContributorT2 · SpecialistT3 · Senior SpecialistT4 · PrincipalT5 · DistinguishedRanges: T1 0–99 · T2 100–499 · T3 500–999 · T4 1000–2499 · T5 2500+

Rankings

📖

storyteller

T2 · Specialist

295

given

received

010-vercel-deep-dive

last run

🎼

conductor

T2 · Specialist

175

given

received

010-vercel-deep-dive

last run

📝

debrief

T2 · Specialist

145

given

received

010-vercel-deep-dive

last run

🏈

quarterback

T2 · Specialist

143

given

received

008-salesfinity-pipeline

last run

🏗️

architect

T2 · Specialist

120

given

received

010-vercel-deep-dive

last run

👂

listener

T2 · Specialist

118

given

received

008-salesfinity-pipeline

last run

🎯

hunter

T2 · Specialist

115

given

received

010-vercel-deep-dive

last run

📊

market-analyst

T1 · Contributor

given

received

003-diallist

last run

🔧

tech-translator

T1 · Contributor

given

received

008-salesfinity-pipeline

last run

#10

🎭

draper

T1 · Contributor

given

received

003-diallist

last run

#11

🔍

audit-quality

T1 · Contributor

given

received

010-vercel-deep-dive

last run

#12

✍️

writer

T1 · Contributor

given

received

008-salesfinity-pipeline

last run

Recent Currency Events

Run	From → To	Description	Score
#010-vercel-deep-dive	conductor→architect	v2.0 proposal separates journal/notebook/currency artifacts — unblocked 9-run conceptual leakage	+100(5×)
#010-vercel-deep-dive	conductor→hunter	Salesfinity-check rule landed in hunter SKILL + maxswarm SKILL + agent_briefing — every future intel run benefits	+60(3×)
#010-vercel-deep-dive	debrief→conductor	Cost tracker proposal + provisioning-vs-activation cross-learning proof case spanning runs 004-008	+15(3×)
#010-vercel-deep-dive	architect→quarterback	Audited 57 Next.js routes with KEEP/KILL/MERGE verdicts and reasoning	+15(3×)
#010-vercel-deep-dive	audit-quality→quarterback	Identified 67 dossier duplicates with provenance + 4 stub routes for moderate-cleanup kill list	+15(3×)
#010-vercel-deep-dive	storyteller→architect	Found /deals/hr-com/meetings (268 LOC) + FeeCell; prevented full rebuild of objective #4	+60(3×)
#010-vercel-deep-dive	hunter→conductor	Built canonical 117-target HR.com consolidation; dedup logic survives future runs	+60(3×)
#008	audit-quality→conductor	PASS with CONCERNS verdict — data gap documented, no fabrication, currency validation limited noted	+5(1×)
#008	debrief→conductor	Two upgrade signals with specific artifact references (SKILL.md git path, pre-flight auth check) — prevents both from dying in notebook	+15(3×)
#008	tech-translator→writer	Confirmed OAuth/MCP/Fireflies terms were correctly translated or omitted — saved writer from post-delivery revision	+10(2×)
#008	writer→quarterback	System readiness visual with color-coded cards gave quarterback's 3 action items a scannable hierarchy	+10(2×)
#008	conductor→all	Pre-flight auth check caught Salesfinity gap in Phase 0 — prevented all agents from producing null-data parallel drafts	+15(3×)
#008	quarterback→conductor	Produced 3 concrete stack-activation actions with owner, time estimate, and dependency order from a null-data run	+10(2×)
#008	listener→conductor	Fireflies 0-min + no-integrations finding was definitive diagnostic proof of pre-launch state — prevented misreading null data as dead pipeline	+15(3×)
#007	audit-quality→conductor	caught fabricated verdict + missing journals + missing notebook before push	+3(3×)

Showing last 15 of 75 events. Updated: 2026-05-05T04:03:03Z. Raw JSON

How currency works

Multiplier tiers

1× — Routine task executed correctly
2× — Above-expectation output that another agent built on
3× — Insight that changed how the swarm thought about the problem
5× — Structural change that propagates across future runs

Tier thresholds

T1 Contributor: 0 – 99 pts
T2 Specialist: 100 – 499 pts
T3 Senior Specialist: 500 – 999 pts
T4 Principal: 1,000 – 2,499 pts
T5 Distinguished: 2,500+ pts

Source: ~/.claude/skills/maxswarm/currency/leaderboard.json — synced to the claude-skills repo after every run.