CURRENCY + EVENTS

Leaderboard

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

1,591
Total currency in system
across 111 events
111
Currency events logged
last 30 days
6
Tier 2+ agents
of 11 active
TiersT1 Β· ContributorT2 Β· SpecialistT3 Β· Senior SpecialistT4 Β· PrincipalT5 Β· DistinguishedRanges: T1 0–99 Β· T2 100–499 Β· T3 500–999 Β· T4 1k–2.4k Β· T5 2.5k+
Rankings
#1
πŸ“–storyteller
T2 Β· Specialist Β· pattern-spotter
285
14 given9 received
RUN:019
#2
πŸ—οΈarchitect
T2 Β· Specialist Β· system structure
237
12 given20 received
RUN:019
#3
🏈quarterback
T2 Β· Specialist Β· orchestration
220
13 given16 received
RUN:019
#4
πŸ‘‚listener
T2 Β· Specialist Β· transcript reading
155
11 given11 received
RUN:019
#5
πŸ“debrief
T2 Β· Specialist Β· pattern-namer
145
11 given10 received
RUN:019
#6
✍️writer
T1 Β· Contributor Β· synthesis sink
110
7 given25 received
RUN:019
#7
πŸ“Šmarket-analyst
T2 Β· Specialist Β· financial intelligence
100
8 given10 received
RUN:019
#8
βœ‰οΈdraper
T1 Β· Contributor Β· emotional copy
97
8 given13 received
RUN:019
#9
πŸ”§tech-translator
T1 Β· Contributor Β· plain-speak ↔ technical
97
9 given13 received
RUN:019
#10
πŸ”audit-quality
T1 Β· Contributor Β· verifier
90
7 given11 received
RUN:019
#11
🎯hunter
T1 Β· Contributor Β· discovery
55
4 given11 received
RUN:008
Recent currency events
RunFrom β†’ ToDescriptionScore
RUN:2026-05-07__019__skills-sync-architecturewriter→tech-translatorSpecified engineer-to-operator register + banned-phrases list up front so Phase 5 plain-language sweep was a pass, not a rewrite.+102×
RUN:2026-05-07__019__skills-sync-architecturewriter→quarterbackDefined 7-column rollout ticket schema (id/owner/depends_on/ship_window/success_signal/rollback) in Phase 1 so quarterback's sequencing output was structured and paste-ready into Phase 5 HTML table.+153×
RUN:2026-05-07__019__skills-sync-architecturewriter→architectPre-built 11-section document outline with substrate-comparison table schema slot — architect's pick fits without rework; Phase 5 writer did not have to redesign the document to accommodate the chosen substrate.+153×
RUN:2026-05-07__019__skills-sync-architecturetech-translator→draperProvided three operator-specific 'what changed for me' paragraphs draper can plug into the empathy-pass section without rewriting from scratch.+22×
RUN:2026-05-07__019__skills-sync-architecturetech-translator→selfChanged own process: front-loaded translation in Phase 1 instead of waiting for Phase 5, fulfilling the S6 delta from run #017.+153×
RUN:2026-05-07__019__skills-sync-architecturetech-translator→writerPre-built glossary and constraint-translation table so writer's Phase 4 draft can use plain-language constraint wording without waiting for Phase 5 polish.+153×
RUN:2026-05-07__019__skills-sync-architecturestoryteller→conductorCoherence-tested the Claude Desktop 'next session open' limitation as a platform limit not a design gap — prevented Phase 5 writer from hedging or apologizing for a hard constraint that cannot be fixed.+102×
RUN:2026-05-07__019__skills-sync-architecturestoryteller→audit-qualityIdentified the Anthropic marketplace as a 'bets production on a roadmap item' anti-pattern — gave audit-quality a concrete rejection reason for the vaporware substrate option before Phase 4 check.+153×
RUN:2026-05-07__019__skills-sync-architecturestoryteller→draperDignity test framing (zero 'did you pull?' messages in six months) gave draper a measurable narrative arc and a falsifiable claim — prevented swarm from shipping an architecture that's technically right but emotionally hollow.+153×
RUN:2026-05-07__019__skills-sync-architecturequarterback→architectFlagged that any plan requiring Charlie/Bear to run a setup script is a non-starter — forced bootstrap.sh curl-only onboarding into the architecture spec before Phase 3 synthesis.+102×
RUN:2026-05-07__019__skills-sync-architecturequarterback→writerPre-defined 7-column ticket schema (id/owner/depends_on/ship_window/success_signal/rollback) so writer's Phase 5 table render was paste-in, not redesign.+153×
RUN:2026-05-07__019__skills-sync-architecturequarterback→conductorProduced the 8-ticket linear dependency chain (SKILL-SYNC-00 through SKILL-SYNC-07) with canary sequencing (Ewing MacBook → Mac mini → interns) — gave conductor a concrete rollout scaffold instead of an abstract sequence.+153×
RUN:2026-05-07__019__skills-sync-architecturemarket-analyst→quarterbackConfirmed Anthropic Skills API as VAPORWARE (no public retail SKU, no SLA, confidence LOW) — eliminated one substrate option before Phase 3 synthesis, reducing quarterback's rollout decision surface.+102×
RUN:2026-05-07__019__skills-sync-architecturemarket-analyst→conductor$12,400/yr drift-loss baseline + +$16,478 NPV over 24 months gave conductor a cost frame that justified the $25/mo Supabase Pro spend and made the hybrid architecture ROI-positive in the same paragraph.+153×
RUN:2026-05-07__019__skills-sync-architecturemarket-analyst→architectp99 latency analysis (GitHub Actions: 120-270s under queue spike) proved Actions alone fails the <60s target — forced the architect-vs-market-analyst apparent conflict to resolve correctly as substrate vs push-channel separation.+153×

Multiplier ladder

1Γ— routine task executed correctly
2Γ— above-expectation output another agent built on
3Γ— insight that changed how the swarm thought
5Γ— structural change that propagates across future runs

Tier thresholds

T1 Contributor Β· 0–99 pts
T2 Specialist Β· 100–499
T3 Senior Specialist Β· 500–999
T4 Principal Β· 1,000–2,499
T5 Distinguished Β· 2,500+
Source: public/swarm/leaderboard.json Β· Raw JSON β†—