Today
Land and start — what needs doing now, refreshed live as work lands.
Mission control
One operating surface for a fleet of agents working in tandem — opportunity intake, security scanning, compliance mapping, and capture drafting — plus the programs, tasks, and risks they feed. Every number below is a doorway: click it to see the underlying items and jump to where you work them.
Opportunities →
Programs →
Security scanning →
Compliance →
AI spend · month to date
$0.00 / $200
Today: $0.00
Used: 0%
Reads token_ledger · live, updates after each run
Platform health
Live endpoint checks · auto-refreshes every 5 minutes
Cockpit · conclavesec.com (Cloudflare Pages)checking…
Supabase (auth · database · realtime)checking…
OIA Droplet · 143.244.176.23 (Python/cron)checking…
State of play
Live snapshot · refreshes on every change
Growth ledger
The metacognitive loop, measured: gaps closed and framing shifts (not volume), plus the open falsifiable bets settling against the live CISA KEV catalog.
Programs pipeline
Grant, contract, and bounty pursuits by stage. Numbers reflect weighted EV (probability × award).
Top open risks
Highest residual scores among active risks.
Today's check-ins
What everyone reported today.
Top open items
Critical and high-priority tasks not yet complete.
Opportunity Intelligence
Daily ranked output from the Opportunity Intake Agent. SAM.gov, SBIR.gov, and bug bounty platforms scored against the Conclave rubric each morning at 06:00.
Awaiting latest OIA run
Droplet 143.244.176.23 · operational since 2026-05-24 · Mon–Fri 06:00 UTC
OIA stack — operational as of 2026-05-24
Federal registrations active
UEI: PDXKRBSZ6ZX5 · EIN: 42-2457630 · SAM.gov registered · CAGE: JRQF5
OIA pipeline running
143.244.176.23 · 11 sources · LangGraph · python3 -m agents.intake_agent
Environment configured
SAM.gov API · Grants.gov API · DigitalOcean Postgres · Supabase service key
Cron schedule installed
Mon–Fri 06:00 UTC daily (~$0.20) · Sunday 06:00 UTC full run (~$0.70) · /etc/cron.d/conclave-oia
Cockpit feed live
push_to_cockpit node writes pursue/maybe opportunities to oia_opportunities after each run
Langfuse tracing active
Every run traced at us.cloud.langfuse.com · token usage + cost tracked · grant evidence recording
Source registry
Live — derived from what each source has actually produced in oia_opportunities, not a hand-kept list.
Loading sources…
Pending review
Opportunities the OIA wants to land in the cockpit. Promote to copy into the feed, Reject to drop. Empty until the OIA stages an opportunity for review.
Opportunity feed
Ranked by rubric score (higher = better fit). Click Pursue to graduate an opportunity into a project.
Tasks
One source of truth — these tasks roll up live into the Today snapshot and each project workspace. Click a task to edit; check it to advance status. Realtime-synced across sessions.
Programs pipeline
Grants, contracts, bounties, and IDIQ vehicles in pursuit. Click any card to edit.
Loading project workspace…
Loading project workspace…
RSI Submissions
Dual-track submission plan: Track A = academic credibility (arXiv → AAAI → USENIX → IEEE S&P), Track B = federal wealth (NSF SBIR → AFWERX → DARPA I2O). NV006 is in-flight and urgent.
Loading…
Roadmap · Wealth & Credibility
Dual-track: Wealth = federal funding (NV006 → NSF SBIR → GOEO → AFWERX → DARPA I2O) and Credibility = academic + registration (arXiv → AAAI → USENIX → IEEE S&P). Total potential: $2.5M–$13M.
Loading…
Sentinel Incidents
What the sentinel overseer flagged — cost spikes, budget breaches, purpose violations, and rubric drift — from public.sentinel_incidents. When an incident paused an agent, Resume clears the pause flag and resolves the incident in one step (via the sentinel_resume RPC), writing a session_log entry.
Loading incidents…
Risk register
Sorted by residual score, highest first. Inherent = Likelihood × Impact before mitigation; Residual = after mitigation. Click any row to edit.
| ID | Risk | Inherent | Residual | Status | Mitigation |
|---|
Compliance coverage
How this connects
The scan engine tags every code finding with a CWE; each CWE maps to the controls it implicates across the frameworks our federal and commercial clients require. This panel rolls that mapping up per framework. It is driven by the same engine as Security Scanning — advisory framing to focus remediation, not a formal attestation.
See the scanners →
Control coverage across the frameworks our federal and commercial clients
care about, driven by the scan engine. Every code finding carries a CWE;
each CWE maps to the controls it implicates. Advisory framing to
focus remediation — not a formal attestation.
Auto-satisfied — clean scan
Gap — findings present
Needs evidence
Human-attested
Process-only
Loading control register…
Security scanning
How this connects
The Conclave application-security engine — one framework, many analyzers, one report — covering source code, dependencies, secrets, infrastructure, and APIs, with SBOM supply-chain reasoning on top. Its findings are what drive the Compliance coverage map, and it is the core capability behind Track A · Sentinel.
See control coverage →Track A · Sentinel →
The Conclave application-security engine — one pluggable framework, many
analyzers, one unified report. Covers the full application gamut across
layers: source code, dependencies, secrets, infrastructure, and APIs.
Independently calibrated against third-party vulnerable applications.
Analyzer coverage
Each analyzer targets a distinct layer of the application surface.
Supply-chain reasoning (SBOM)
Beyond "is it vulnerable?" — to "is it reachable, exploitable, and does it matter?"
Calibration scorecard
Scored against an independent corpus of deliberately-vulnerable third-party apps. We report category recall — not a fabricated accuracy number.
Architecture & asset inventory
Live infrastructure topology and asset inventory. Click any box on the diagram or any row in the table to open the detail drawer. Edits are admin-only; retire-via-status rather than delete.
Platform topology
Click any box for details · status reflected by border styleAsset inventory
| Name | Category | Provider | Status | $/mo | One-time | Description |
|---|
Today's check-in
File one check-in per session. If you've already filed today, you'll edit your existing entry.
History
All check-ins, newest first.
Canonical foundation files
Links to the source-of-truth artifacts in C:\ClaudeCode\ConclaveSecurity_00\. Open the path on the machine where this dashboard is being used.
Deliverables
Generated, downloadable artifacts — whitepapers, prior-art reports, patent drafts, licensing packages. Produced and validated by the self-build harness (tools/selfbuild/); each renders to a self-contained, print-to-PDF document.
Loading deliverables…
Capability Ledger PRODUCTION · founded 2026-06-11
Production public.ledger (internal/confidential) was founded 2026-06-11 by a 2-of-2 sovereign ceremony (ADR-022/028) — height 0 fingerprint c90123b0…e8f504, externally anchored to Bitcoin block 953174 + OpenTimestamps. Governance is live: the Authority Triad keys are registered (h1), quorum enforcement is active from h2, and the first capability claims are sealed (h3). The preserved proof-of-concept chain (public.ledger_poc, height 16) remains as reduction-to-practice history. action_log is the activity stream.
Scope (labeled honestly): hash-linked core + deterministic replay + the full Authority Triad now enforced on production — every block above the sealed enforce_from boundary (h2) is checked for separation-of-duties quorum (builder proposes; agent + constitutional validator + sovereign approve per tier). Append-only is enforced at the database (UPDATE/DELETE rejected) and the chain is read-only from the client (writes are a sovereign/triad ceremony act via the service role). Hard finality (epoch Merkle checkpoints) is the next layer. “Verify” below recomputes every block’s content-hash and back-link in your browser — the §4.3 replay acid test against what’s actually stored.
Loading ledger…
Decisions
Append-only record of material decisions. Edit the decision-log.md file in tandem.
Session log
Append-only audit trail. Newest first.
| When | Who | Action | Details |
|---|
Automatable backlog
Structured inbox for work that should become agent-driven. Pipeline: submitted → triaged → scoped → building → shipped. Format spec: 03_AUTOMATABLE_BACKLOG_FORMAT.md.
Loading backlog…
AI spend
Token + API spend by agent run. Reads token_ledger — updated after each pipeline execution.
Month to date spend
$0.00
of $200 cap
Today
$0.00
this calendar day
Runs MTD
0
pipeline executions
Avg/run
$0.00
triage + qualify
Spend by workstream · MTD
Recent OIA runs
| Date | Run ID | Triage (Haiku) | Qualify+Reflect (Sonnet) | Total cost | Tokens in | Tokens out | Pursued |
|---|---|---|---|---|---|---|---|
| No runs recorded yet — cost data appears after first OIA execution. | |||||||
Agents & Ops
The live hive — agents, supporting runtime, and the quality gate — from public.assets, with status, last activity, and metered model spend. Manifest-ready: merges agents/plugins/*/manifest.yaml as plugins ship.
Loading hive…
Settings
Workstream names, cost-pool config, your profile, and live integration health.
Loading settings…
Advisory desk
How this connects
A standing executive board for the firm. Five specialist advisors — CFO, COO, CTO, CISO, CRO — each read the live org (risks, backlog, spend, delivery) and take a stance; the Chief of Staff synthesizes them into one recommendation. The board's calls feed the Build queue, where you approve the work that ships. Briefs are versioned in
discovery_proposals so every recommendation is auditable.
Go to Build queue →
See top risks →
Loading board…
Brief history—
Every board session and discovery proposal, newest first. Open one for the full advisor breakdown and write-up.
Loading briefs…
Where board recommendations become shipped work. Approve proposed tasks, run a build shift, then review and verify the output before it counts.
Queue — Approved & Staged
0
Loading…
Shift Results — Review & Accept
0
Loading…
Top Risks
Loading…
Sprint History
Loading sprints…
Feed — all cockpit updates in one place
Loading intelligence…
Trigger an autonomous agent workflow. Each action runs asynchronously — results surface in Pipeline and Intelligence.
Loading recent triggers…