Loading usage feed…
How to read this.
A percentage appears only when the provider genuinely reports one.
UNSUPPORTED means the provider exposes no
programmatic usage surface; UNAVAILABLE means the surface exists but isn't reachable/authenticated right now;
STALE means the last successful refresh is over 15 minutes old. Feed:
hq.provider-usage/v1 via /mc/provider-usage, produced by the System collector — HQ reads a file and
holds no tokens, cookies, or keys.
👑 1 Percenter Mafia
⠿ agents · ☰ teams · edit names/subtitles
Family
5 agents
Family
KH
Available
medium
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
medium
reasoning
Fallback 1
None
—
Configure
Fallback 1
Family • Default
UB
Available
high
Model
Grok 4.5
xai-oauth
Effort
high
reasoning
Fallback 1
GPT-5.6 Luna (Codex)
medium
Configure
Fallback 1
Family
CS
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Family
EN
Available
low
Model
Claude Opus 5
anthropic
Effort
low
reasoning
Fallback 1
GPT-5.6 Sol (Codex)
low
Configure
Fallback 1
Family
CT
Available
low
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
low
reasoning
Fallback 1
None
—
Configure
Fallback 1
Intel
6 agents
Intel
IC
Available
high
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Intel
RS
Available
high
Model
Grok 4.5
xai-oauth
Effort
high
reasoning
Fallback 1
GPT-5.6 Luna (Codex)
max
Configure
Fallback 1
Intel
IN
Available
low
Model
Grok 4.3
xai-oauth
Effort
low
reasoning
Fallback 1
GPT-5.6 Luna (Codex)
medium
Configure
Fallback 1
Intel
HK
Available
xhigh
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
xhigh
reasoning
Fallback 1
None
—
Configure
Fallback 1
Intel
D
Available
low
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
low
reasoning
Fallback 1
None
—
Configure
Fallback 1
Intel
E
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
8 agents
Developers
DC
Available
high
Model
Claude Opus 5
anthropic
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
AR
Available
high
Model
Claude Fable 5
anthropic
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
RV
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
MR
Available
max
Model
GPT-5.6 Terra (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
CD
Available
xhigh
Model
GPT-5.6 Terra (Codex)
openai-codex
Effort
xhigh
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
IM
Available
high
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
DR
Available
high
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Developers
S
Available
max
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Financial
4 agents
Financial
FC
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Financial
AC
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Financial
AN
Available
max
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Financial
BK
Available
xhigh
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
xhigh
reasoning
Fallback 1
None
—
Configure
Fallback 1
Product
2 agents
Product • Default
D
Available
high
Model
Grok 4.5
xai-oauth
Effort
high
reasoning
Fallback 1
GPT-5.6 Luna (Codex)
medium
Configure
Fallback 1
Product
SF
Available
default
Model
unknown
provider n/a
Effort
default
reasoning
Fallback 1
None
—
Configure
Fallback 1
Marketing
2 agents
Marketing
WR
Available
high
Model
GPT-5.6 Terra (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Marketing
PB
Available
max
Model
GPT-5.6 Terra (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Legal
3 agents
Legal
LC
Available
max
Model
Claude Fable 5
anthropic
Effort
max
reasoning
Fallback 1
GPT-5.6 Sol (Codex)
max
Configure
Fallback 1
Legal
CO
Available
high
Model
Grok 4.5
xai-oauth
Effort
high
reasoning
Fallback 1
GPT-5.6 Sol (Codex)
high
Configure
Fallback 1
Legal
PL
Available
max
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Wellness
2 agents
Wellness
WC
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Wellness
TR
Available
max
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
max
reasoning
Fallback 1
None
—
Configure
Fallback 1
Specialists
4 agents
Specialists
AS
Available
xhigh
Model
GPT-5.6 Luna (Codex)
openai-codex
Effort
xhigh
reasoning
Fallback 1
None
—
Configure
Fallback 1
Specialists
SN
Available
high
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
GPT-5.6 Sol (Codex)
high
Configure
Fallback 1
Specialists
RB
Available
high
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
GPT-5.6 Sol (Codex)
high
Configure
Fallback 1
Specialists
O
Available
high
Model
GPT-5.6 Sol (Codex)
openai-codex
Effort
high
reasoning
Fallback 1
None
—
Configure
Fallback 1
Command map / design preview
Orchestration see the machine behind the cards
One visual surface for the live profile roster, team leads, worker branches, cross-team handoffs, and the judgment points that close a chain. Rearrange the draft with drag-and-drop, then tune how much the lead owns before anything would be promoted to runtime.
Reading live roster…
Profiles
—
live Hermes roster
Lead seats
9
intake + close authority
Graph families
8
team graphs on disk
Draft state
Clean
local design changes only
Completion p95
—
Waiting for telemetry
Queue p95
—
Created → worker start
Failure / recovery
—
Crash, timeout, block, reclaim
Estimated cost
—
Usage records in window
Baseline chain P50
—
Observed work + queue
Draft projection P50
—
No staged changes
Projected delta
—
Heuristic, not a runtime promise
Confidence
—
Stage coverage required
Routing graph
Solid lines are normal routing. Dashed lines are judgment, conditional escalation, or cross-lead handoff.
Drag nodes · click to inspect
Prototype controls: drag any node · drop roster chips onto a lead · select a lead to tune involvement
normal route
judgment checkpoint
cross-lead handoff
Node position and settings are staged locally
Bottleneck watch
Profiles ranked by tail latency, queue pressure, and recovery rate. Click a row to inspect that node.
Loading bottleneck data…
Slow traces
A trace is evidence, not a verdict. Open the node inspector for cohort context.
Loading trace data…
Stage evidence
Only structured stage-start / stage-complete events appear here. Missing rows are not inferred from profile timing.
Waiting for explicit stage events…
Draft impact
Only changes with an explicit heuristic are scored. Everything else stays visibly unmodeled.
Switch to Design and stage a change to see its projected impact.
Simulation assumptionslow confidence
The simulator is a design aid. It does not claim to predict runtime behavior or quality.
Waiting for observed telemetry.
Active chain preview
The selected view translated into a sequence a lead could actually operate.
Org topology
Team graph library
Choose a lane to focus the map. Each tile is also a drop target for an agent chip.
8 graph families
Design mode only. Dragging, sliders, toggles, and roster drops create a local draft in this browser. Nothing is written to Hermes routing, Kanban, profile identity, or runtime config yet. The disabled apply button is deliberate; this first pass is for Don to react to the shape of the control surface.
✅ Tasks
Ideas1
PARK: retire legacy underboss profile (OpenClaw default leftover)
Waiting7
P0 STORAGE PASS 2: recover actual GiBs through verified bulk offload
RECOVER: LIVE: install verified state.db candidate and prove default surfaces
LIVE: install verified state.db candidate and prove default surfaces
RECOVER: restore state.db and fail-close live malformed repair
DEV: 1% Heist universal exit — middle-ground logic redesign
DEV-BUILD: 1% Heist v4 G6d — evidenced non-reversion proof on Chart 2
SYSTEM: restore Claude Fable usage in HQ Subscription Limits
Active0
Blocked2
OPERATE: arm and resume existing Bookkeeper PASS2 through trusted scoped permits
EXECUTE: storage pass 2 offload from verified 37.9 GB inventory
Done290
RECOVER: SYSTEM FIX: hard-stop post-artifact GUI/CDP no-progress
INTEL CORRECTION: capability-first 5/10 writer contracts
SYSTEM FIX: hard-stop post-artifact GUI/CDP no-progress
DIAGNOSE: run 2200 repeated CDP no-progress loop
P0 SYSTEM: prevent repeated browser/CDP no-progress loops
REVIEW: scoped kanban_rearm authority + deny gates (seat-corrected re-arm of t_d79ae25a)
DECISION: TEST AUTHOR FIX3: executable RED artifacts for 18 kanban_rearm probes
DECISION: live canary — exercise binding Consigliere ROUTE authority once
TEST AUTHOR: 18 behavioral RED probes for kanban_rearm authority
TEST AUTHOR FIX3: executable RED artifacts for 18 kanban_rearm probes
DECISION: P0 VERIFY: System + Intel + Development team-chain audit
DECISION: JUDGE FIX1: replace false-green kanban_rearm placeholders with 18 executable RED probes
DECISION: FIX: claude_cli runtime drops the entire Hermes system prompt (18 profiles blind)
JUDGE: RED-baseline scoped kanban_rearm authority tests
P0 DEV LEAD: implement scoped kanban_rearm authority + deny gates
S1 FINDINGS: map scoped re-arm authority and t_43113269 collision
F9 JUDGE (FINAL): re-judge task_comments lineage channel + full regression (checks 1-9, D1-D5)
F9 FIX (FINAL): add task_comments lineage channel + drop 2 bad rejected_citations
DECISION: fleet lead-routing audit — authorise final bounded re-arm, or ship with documented residual?
FIX: swarm policy-skill pins hard-break `hermes kanban swarm` under the new fail-closed gate
F4 FIX: assignment-history rule + reconcile body exception; regenerate fleet lead-routing artifacts
INTEL F5: re-judge the corrected fleet lead-routing product (fix-loop 1 verdict)
INTEL F4: fix fleet lead-routing product — assignment history, body exceptions, number-block assertions
FIX1 DETECTIVE: correct System roster prompt-layer audit after Enforcer reject
P0 DEV: HQ canonical identity/role propagation + rollback round-trip
INTEL F3: adversarial judge — break and verify the fleet lead-routing product
INTEL F2: fleet lead-routing classification engine + metrics (108-row frozen cohort)
INTEL F1: cause signatures for lead solo-execution (VERIFIED vs UNVERIFIED at spawn)
PEER INTEL: fleet-wide true lead fan-out + card classification
DEEP AUDIT: prompt-injection loss + graph-membership deadlock + intake truncation
RETRY1 DETECTIVE: true lead fan-out rate + 108-card routing classification
FIX: claude_cli runtime drops the entire Hermes system prompt (18 profiles blind)
P0 VERIFY: System + Intel + Development team-chain audit
P0 INTEL: audit Investigator/Analyst/Researcher/Bookkeeper/Informer chain
P0 DEV: Kanban lead prompt/skill injection + routing enforcement audit
P0 AUDIT: fleet team chains/loops/graphs + lead-routing enforcement
SYSTEM: clear 3 stale/false Blocked cards; preserve PARK/device holds
REVIEW: claude_cli system-prompt fix (t_2caa1ffb) — core path-class, deep review
ARCH FIX1: owned-child wait_on must auto-resume, not deadlock
ARCH: provenance fan-out + stale-run fencing + skill/intake integrity
ARCH: canonical HQ agent identity-write transaction + stable profile-id policy
IMPL: kanban_complete acceptance parser — stop demanding read-only Inputs as deliverables
SYSTEM REPAIR: kanban_complete mis-resolves acceptance input path
FIX Gap C: lead-only doctrine on junior seats + a distribution rule keyed on routing_role
DETECTIVE: exact System roster tools/skills/runtime-injection fitness
INTEL AUDIT I1: Informer pull-seat ownership proof + scoped capture canary
INTEL AUDIT B1: Bookkeeper librarian-role consistency + live sort canary
INTEL AUDIT A1: quantitative routing history for the 5 Intel seats
INTEL AUDIT R1: doctrine corpus dossier for the 5 Intel seats
FIX (M3): ROUTING_DOCTRINE.md is stale and contradicts all three lane maps — 0/5 sub-checks pass
FIX (W1+D1): structured wait_on absent from all 71 carriers; 34 SOULs say the recovery watchers are disabled while both are LIVE
FIX (M1+M2): dev lane map header + Full-path seat order — REJECTED by final routing contract
ARCH: route-authority metadata without team-lead semantics (re-scope after t_ed040823 exhaust)
RECOVER: ARCH: route-authority metadata without team-lead semantics
RECOVER: ARCH: route-authority metadata without team-lead semantics
ARCH: route-authority metadata without team-lead semantics
GAP H: make doctrine tell the truth about the two stopped recovery daemons
SYSTEM S2 FIX: isolate verifier + reconcile Grok live contract
IMPL: persist drag order for every HQ Usage card
N2 IMPLEMENT: canonical-route guardian in hq-reliability-check.js + disarm phone-access-keeper 18789 landmine
N1 INVESTIGATE (read-only): who rewrote Funnel / → retired :18789 on 2026-07-31
SYSTEM: prevent HQ Funnel drift back to retired :18789
IMPL: Kanban lead control plane — provenance completion gate, structured wait_on, run fence, intake floor
N3 FORENSICS: attribute the 2026-07-31 HQ Funnel rewrite to :18789 (deep dig after fast pass failed)
N4 VERIFY: evidence-grade audit of HQ Funnel attribution residual
IMPL: routing_role `authority` — deliver kanban-team-routing to Consigliere without lead semantics
SYSTEM/S1b: instrument the real Grok client — intercept CLI /usage + authenticated grok.com quota surface
REPAIR: authorized PARTIAL disposition for loop-exhausted lead card t_f1120aba
PREFLIGHT: routing_role authority implementation
DEV: routing_role enum has no value for a route-authorised non-lead seat (consigliere gets no routing skill)
SYSTEM/S1: find a real Grok/SuperGrok subscription-usage surface (evidence only, no implementation)
DEV: Usage-card drag/drop persistence + Grok usage integration
SYSTEM: obtain real Grok/SuperGrok subscription usage for HQ
FIX-2: bridge RUN-only malformed worker through _touch_activity
DECISION: REARM-ONCE VERIFY: Architect design match after FIX-1
Verify design match for b47e18bf5 after FIX-1
IMPL: stale-worker write fence + prior-worker spawn guard
ARCH: fence stale Kanban worker writes and orphan redispatch
REVIEW: stale-worker write fence and prior-worker spawn guard
RECOVER: AUDIT: F2 PreToolUse choke point — production G3 + rollback re-audit
VERIFY: final Architect design match after C4 FIX-2
AUDIT: F2 PreToolUse choke point — production G3 + rollback re-audit
DECISION: HARNESS DEFECT: kanban worker outlives its own terminal board call → two agent loops on one card
HARNESS DEFECT: kanban worker outlives its own terminal board call → two agent loops on one card
Doctrine delivery gaps: profile AGENTS.md never loaded + no role-derived skills injection
CONTAIN: restore scoped permits default-OFF after S6 precondition FAIL
EXTERMINATE: bind Kanban progress guard to real log semantics
ADMIN: commit Legal authority read-back
DECISION: preserve Consigliere routing under authority enforcement
CORRECT: remove final Option A prompt bypass grants
ADMIN: align System doctrine with live progress circuit breakers
CORRECT: tighten Kanban progress signals and typed exemptions
BREAK: authority-split bypass and activation review
VERIFY: Option A code with uncontaminated binary harness
SYSTEM RESTORE: re-arm existing PASS2 Developer-Capo lead after archival race
SENIOR BUILD: close root directory lstat-open race
SENIOR REVIEW: S1-S8 final root-identity and permit gate
ALIGN: Option A managed prompt semantics
BUILD: fresh Option A A08 activation judge
SENIOR REVIEW: PASS2 cap and full live-boundary S1-S8
RECOVER: SENIOR BUILD: close PASS2 cap and full live-inventory gaps
MID REVIEW: PASS2 manifest policy and live-readback M1-M8
RECOVER: MID BUILD: close PASS2 manifest policy and live-readback defects
READBACK: OG Option A owner approval
READBACK: Legal lead Option A owner approval
READBACK: Investigator Option A owner approval
READBACK: Developer-Capo Option A owner approval
READBACK: CFO Option A owner approval
RESET JUNIOR REVIEW: safe exact PASS2 manifest-to-contract gates
ROUTE: resume existing PASS2 S6 Development lead for one bounded corrective ladder
RECOVER: JUNIOR BUILD: safe exact PASS2 manifest-to-contract preparation
RECOVER: offline state.db and restore scoped-permit controller
MID BUILD: fix directory identity, predecessor run, board audit routing
FIX: remove stale Option A direct-junior prompt grants
FINDINGS: map Option A authority remediation surface
MID REVIEW: M1-M8 scoped-permit correction gate
DETECT: exact malformed field in live scoped-terminal preflight draft
FIX: terminalize malformed scoped-permit consume authority
FINDINGS: inventory Option A gateway code residency
READBACK: Default Option A managed-prompt source plan
READBACK: Consigliere Option A prompt plan
JUNIOR RESET REVIEW RECOVERY: reverse cache R1-R7
MID REVIEW: M1-M8 exact consumed-wait receipt gate
MID BUILD: require exact consumed controller-wait receipt
JUNIOR RESET BUILD: adopt scoped-permit bridge with corrected scope
JUNIOR BUILD: bind API request profile into controller session
MID REVIEW RECOVERY: scoped-controller cache M1-M9
JUNIOR RESET REVIEW: Q1-Q8 exact controller-wait bypass gate
JUNIOR RESET BUILD: close exact controller-wait bypasses
JUNIOR RESET SENTINEL: corrected R1-R8 operational permit gate
SYSTEM POST-CLOSE: build fail-closed controller activation script
SYSTEM: close non-authorizing contract-preflight digest gap
JUNIOR REVIEW: API request profile binding seam R1-R8
DEVELOPMENT RESIDUAL: bind API request profile into scoped controller session
RECOVER: MID BUILD: close scoped-controller cache leak and real API seam
JUNIOR BUILD: expose scoped-permit controller tools to API server
P0 SYSTEM: recover recurrent state.db corruption and restore scoped-permit controller
JUNIOR RESET REVIEW: R1-R8 controller-wait exact revision
JUNIOR RESET BUILD: adopt and commit controller-wait candidate
SYSTEM: remediate authority split under Consigliere Option A
READ-BACK: Legal lane authority roster
READ-BACK: Finance lane authority roster
CORRECT: fail closed when HERMES_PROFILE is missing
IMPLEMENT: enforce System-harness versus team-lead graph authority
DIAGNOSE: path-scoped approval choke for authorized Bookkeeper Kanban work
JUNIOR REVIEW: API scoped-permit controller tool exposure G1-G8
DEVELOP: expose scoped-permit controller tools through authenticated API toolset
P0 DEV: add truthful goal-mode controller wait for scoped permit continuation
REVIEW SENIOR: final scoped permit R1-R8 + 38-gate verifier
BUILD SENIOR: close scoped permit consume + terminal boundary
P0 DEV: make scoped permits operational for existing multi-step Bookkeeper PASS2
P0 SYSTEM: operationalize scoped permit for existing Bookkeeper PASS2 card
SECURITY GATE: read-only bypass review of scoped terminal permit fix
REVIEW M5: prepared-ticket capability gates
BUILD M5: bind consume to exact prepared ticket
BUILD J5: authenticated scoped-permit consume channel
P0 SYSTEM: unblock authorized Bookkeeper storage commands safely
REVIEW J5: authenticated consume-channel gates
REVIEW M4: Phase-A freshness regression gates
BUILD M4: recheck scoped permit freshness at Phase A
REVIEW J4: scoped permit Phase-A gates
BUILD J4: scoped permit Phase-A approval preparation
REVIEW: corrected spawn S4 fail-closed gates
BUILD: explicit spawn S4 fail-closed evidence
BUILD J1: strict scoped-permit operation policy
DESIGN: scoped Kanban terminal permit security boundary
JUNIOR SENTINEL FINAL: spawn-channel evidence map
REVIEW JUNIOR RESET: scoped permit arm tool A1-A8
BUILD JUNIOR RESET: finish scoped permit arm tool
MID REVIEW: operation-policy G8 provenance correction
MID FIX: correct operation-policy receipt provenance
REVIEW J1: strict scoped-permit operation policy
BUILD SLICE 1: headless Kanban approval hygiene
P0 DEV EVIDENCE: scoped Kanban terminal permit + headless approval hygiene
RECOVER: EXECUTE: storage pass 2 offload from verified 37.9 GB inventory
JUNIOR RESET BUILD: adopt verified routed-Idea revision
SENTINEL RESET REVIEW: routed-Idea corrected J1-J8
ROUTE: archive invalid Kharon card t_fb4534c0
ROUTE: archive invalid Kharon readback card
CORRECT INSTALL: finish bounded Hermes five-capability adoption
SOLDIER: implement default-OFF Kanban durable-progress guard
DETECTIVE: quantify Kanban no-progress and post-evidence loops
ROUTE: remediate authority split after decision
IMPL: private read-only Ops Attention Center Desktop plugin
BUILD: private Ops Attention Center Desktop plugin
IMPL: remove stray pre-doctype HQ cache line
FIX MOBILE HQ: remove stray dev-three-tier text before doctype
P0 STORAGE: offload verified Hermes archives to private Google Drive
JUNIOR RESET BUILD: freeze isolated Fast Mode acceptance receipt
MID BUILD: isolate routed-Idea transition from legacy cleanup
JUNIOR RESET REVIEW: isolated Fast Mode J1-J8
MID BUILD: persist Fallback 1 effort-only edits
MID REVIEW: routed-Idea transition isolation M1-M8
MID REVIEW: Fallback 1 effort-only persistence
FIX HQ: auto-clear routed Ideas with idempotent tombstone
JUNIOR BUILD: idempotent routed-Idea tombstone auto-clear
JUNIOR SENTINEL FINAL: evidence-map Ops Attention verdict
ADMIN: add contract-quality gate to six Development profiles
SENTINEL REVIEW: routed-Idea tombstone frozen J1-J8
ADMIN: normalize six Development profiles to mandatory paired tiers
CORRECT FINDINGS: selective adoption target and cron gap
FIX: integrated runtime fallback and Kanban protocol evidence
JUNIOR FINAL BUILD: close preserved Ops Attention diff
SYSTEM: stop post-evidence and no-artifact Kanban execution loops
FINDINGS: Hermes selective adoption mechanics
AUTHORIZED SEND: publish reviewed auth seat-belt fix upstream
IMPL: activate configured fallback chain across runtimes (dispatch)
IMPL: activate configured fallback chain across runtimes
P0 FIX: configured model fallbacks must activate across runtimes
MID BUILD: repair Ops Attention exact HQ parity and false-green handling
MID REVIEW: gate exact Ops Attention parity and failure visibility
AMEND CLOSE: Ops semantic parity after Sentinel review
REVIEW JUNIOR: Ops Attention semantic parity
IMPL: semantic Ops Attention parity with HQ
DRAFTER: make Ops Attention semantic counts match HQ
FIX: align Ops Attention Blocked/Waiting with HQ semantic projection
ADOPT: useful Hermes update capabilities
NORMALIZE DEV TEAM: align six profiles to three-tier coder-reviewer ladder
CLOSE: Ideas consolidation and bounded Drafter correction
FIX HQ: persist native Ideas cards across sync and refresh
CORRECT: preserve DON-ONLY/PARK precedence on support leaves
IMPL: stop recursive Kanban support-card cascades
P0 FIX: stop recursive hygiene-on-hygiene Kanban cascades
SYSTEM AUDIT: Hermes update features worth adopting
IMPLEMENT: fail-closed worker residual escalation to owning lead
SYSTEM: enforce worker residual ladder Soldier→lead→Consigliere→Don (no false needs_input)
IMPL: make HQ Log grouped notable changes since last visit
CANARY CODER: Junior Luna/high artifact
TEST AUTHOR: freeze three-tier rollout judges
ARCHITECT DESIGN: three-tier Dev ladder rollout
P0 ROLLOUT: three-tier Dev coder/reviewer ladder
Implement explicit A0 → Associate routing guard and doctrine
SYSTEM: prefer Associate for true A0 micro-fixes (stop Soldier-default underuse)
FIX HQ Log: notable changes since last visit, not completed-task feed
IMPL: finish HQ Log tab checkpoint
IMPL: give reviewer-mid a distinct HQ avatar
HQ: add Log tab for completed change history
HQ: add avatar for reviewer-mid
P1 HARDEN: Kanban stuck lifecycle and automatic recovery
REPAIR 2 FIX-LOOP 1: bind Enter ack reconciliation to the real dedicated callout queue
IMPL REPAIR 2: Discord Enter ack/rewrite terminal integrity
P0 REPAIR 2 ENG: close Discord Enter ack/rewrite loss + green frozen suite
JUNIOR RESET BUILD: adopt Ops TDZ bytes + correct live receipt
P0 VERIFY REPAIR 2: Discord Enter ack/rewrite integrity
SENTINEL RESET REVIEW: Ops TDZ live receipt R1-R8
BUILD JUNIOR: Fix Ops Attention rows TDZ + executable regression
IMPL: Discord Enter publish integrity — kill ghost/cross-play public-enter, make Enter loss loud, offline replay
P0 ENG: repair Discord Enter publish integrity + offline replay
REVIEW JUNIOR: Frozen J1-J8 gate for Ops Attention rows TDZ
P0 VERIFY: Discord Enter publish repair (read-only)
FIX DESKTOP: Ops Attention rows initialization crash
DECISION: SYSTEM FIX2: close remaining org-verdict authority escapes
SYSTEM FIX2: close remaining org-verdict authority escapes
DECISION: FIX S2: t_f79eb3de adversarial rejects — prededup/protected/lead/idempotency
FIX S2: t_f79eb3de adversarial rejects — prededup/protected/lead/idempotency
INVESTIGATE: AP1–AP6 Kanban stuck-pattern guard seams
P0 ROBBER: Discord Enter publish — kill ghost public-enter / expired loss / rewrite race
RECALIBRATE Kharon fill % (too generous) + deep avatar questionnaire
LEAD: Build Kharon as Don-mirror brain (replace you over time) + HQ Don fill tab
DESIGN: Stage-2 goal-driven autonomy (daily plan from Don goals — next leap)
SYSTEM FIX1: close org-verdict authority guard escapes
S1 INSTALL: stuck-pattern create guards + lead-dispose watchdog
SYSTEM: stuck-pattern prevention — unassigned refuse + loop_exhaust→lead CLOSE + harness no-litter
INVESTIGATE: org-verdict authority guard insertion points
JUDGMENT: What is the evolution after graphs — and are we building it?
Obsidian P0-A1: triage the 22 root Untitled*.md files (front-door cleanup only)
SYSTEM ENFORCE: only team leads or Consigliere may hold Judge authority
DECISION: Obsidian P0-A: vault front door — HOME.md index, top-level MOCs, _inbox capture surface, root Untitled triage
Stage-2 2.0b: real dry-run morning plans (you fill ranks) + morning schedule — NO auto-route
Doctrine+watchdog: stop false Waiting (mid-pass needs_input, loop-exhausted recovery, long-card budgets)
Obsidian P0-A: vault front door — HOME.md index, top-level MOCs, _inbox capture surface, root Untitled triage
LEAD: Obsidian vault improvements from TB-obsidian (Don priority)
READ books → prioritized org implementation agenda (all library)
X articles: fix broken removal + prove capture backlog (CDP path failing)
IMPL: class-wide Dev junior evidence-only doctrine + stale-decision supersession
DEV CLOSE: Capo owns verdict; junior seats produce evidence only
Obsidian P0-A2: seven tree MOCs (X Articles, Receipts, Wellness, Betting, Knowledge Banks, Intel, System Optimization)
Obsidian P0-B: retrieval layer — note templates, daily-note capture template, Bases dashboards (orphans/projects/health)
INSTALL: Don org-verdict topology on default + Enforcer loaded surfaces
READ TB-obsidian → implement Obsidian vault improvements (Don priority)
SYSTEM ROLLOUT: load the kanban lead-authority guard into the running fleet (drained gateway + dispatcher restart)
FIND: live org-judge misuse + loaded doctrine surfaces for t_2050e4a8
FIX: kharon_loop never posts again once a pending question goes unanswered (25 days / 50 silent runs)
FIX 1: Stage-2 2.0a Enforcer gate — traversal, ceilings, mirror validation
Kharon vault structure: proper mirror + future judgment stand-in layout
IMPL: Obsidian vault structure from TB-obsidian (hands under Investigator lead)
READ system books → System backlog + team graph/loop/harness upgrades
READ eng books → Dev implementation backlog + graph/workflow deltas
System harness pack: terminal contract + dual-run stand-down + verifier kit
X articles PULL+REMOVAL: fresh drain after CDP health check
📜 Log 0
Loading notable changes…
Loading notable changes…
💰 MRR Gauge
$0
0% of goal
📈 Projections
Current MRR$0
Projected ARR$0
Monthly Growth Needed—
Months to Hit Goal—
📊 6-Month Revenue
👥 Clients
| Name | MRR | Status | Since |
|---|
🍩 MRR Breakdown
🎬 Video Pipeline
📈 Growth Tracker
🌱 Rookie Creator
0
1,000
Current Subscribers
0
Members
—
humans
Next Goal
—
milestone
Progress
—
to next milestone
Grand Goal
1K
by Jan 1, 2027
🏆 Growth Milestones
👥 Members
🔥 Topic Patterns Today
Multi-source signal — same topic hitting across different channels
Loading…
🔴 Breaking Now
Loading…
🟡 Building — Watch
Loading…
🧾 Discord Intel Feed
Loading…
📊 Market
Source pending…
SECTOR ROTATION — Snapshot
⚡ Live Callouts — Heist / Stickup
—
Loading live Heist/Stickup board…
📚 Legacy research (last build —)
Research-build play cards (odte/intraday/swing/leaps/futures). Not the live alert doctrine board above.
Loading legacy research plays…
👁️ Watchlist
Robber's picks · direction + thesis + play
Loading watchlist…
🌊 Catalysts
What's driving everything today
Loading themes…
📈 Daily Performance Tracker
Weekdays only
Correct
—
—
Wrong
—
—
Hit Rate
—
—
Streak
—
—
Loading performance…
₿ Bitcoin Tracker
Price · Dominance · Bias · Key levels
LEVELS:
Loading Bitcoin thesis…
🔭 Crypto Watchlist
Top opportunities · direction + thesis + play
Loading watchlist…
⚡ Crypto Plays
Loading plays…
🏦 DeFi Signals
From Informer pattern detection
Loading DeFi signals from Informer scan…
🎯 ALL Daily Card
LOADING
Loading brain status…
Fill = 60% coverage + 40% Don-verified accuracy — a sector cannot pass 60% until Don confirms Kharon's reads on it.
★ = judgment-load-bearing sector (weighted 2× in overall). Data:
don-brain-status.json, regenerated by the Kharon loop after each interview cycle.
𝕏
X / Twitter
—
📸
Instagram
—
🎵
TikTok
—
▶️
YouTube
—
💼
LinkedIn
—
💡 Idea Backlog
from vault
🔄 Pipeline
by status
💡 Ideas
—
✏️ In Draft
—
📅 Scheduled
—
✅ Published
—
Discord Growth
Members
—
Next goal
—
📋 Content Calendar
Platform coverage this week
🤖 Crew Status
📣 Marketer: Active
✍️ Writer: Active
📊 Analyst: Active
🤝 Closer: Active
Active Clients
0
paying
MRR
$0
monthly recurring
Pipeline
0
leads in progress
MRR Goal
$10K
target
💰 Revenue Progress to $10K MRR
0%
$0$2.5K$5K$7.5K$10K
🤝 Active Clients
🏗️
No clients yet
First client gets added here
📊 Lead Pipeline
🎯 Prospecting
—
📧 Contacted
—
📞 Meeting Booked
—
📋 Proposal Sent
—
✅ Closed / Won
—
Pipeline data from
Capo Brain/Leads/
🎯 Target Verticals
Active focus areas
🏥 Homecare Agencies
Local SEO · Lead gen · Social content · Referral source marketing
🟢 Active
🎩 1 Percenter
Discord growth · Content · Platform presence · Community
🟢 Active
👤 Personal Brands
Thought leadership · Content strategy · Platform growth
🟡 Building
