OSINT-041-probabilistic-forecasting-dashboard-collection

OSINT-041 Probabilistic Forecasting Dashboard - Collection Map

OSINT-041 Probabilistic Forecasting Dashboard - collection workspace. Central topic = the forecast question set / strategic topic the dashboard tracks. Branches = the data-point categories for this forecasting dashboard. Drop each collected datum as a child node; expand it with the source, date accessed, and what it implies for dashboard configuration or live probability updates. Paste raw tool output into the node’s Notes. → feeds deliverable OSINT-041 and re-scores upstream forecasts in OSINT-037, OSINT-038, OSINT-039, and OSINT-040.

00 · Collection Plan - PIRs & EEIs

The questions this collection must answer + the essential elements of information (EEI) for each. Tick as satisfied. → drives dashboard configuration, live forecast scores, signpost status, and escalation routing. Tied to the forecast questions resolved in OSINT-037 §4, the scenario signposts in OSINT-038 §11, the watch items from OSINT-039 §8, and the I&W indicator set in OSINT-040 §1.

KIQ-1 - What is the current calibrated probability for each active forecast question?

• EEI: point probability and range for each active FQ from OSINT-037 §4 / §8
• EEI: ICD 203 band cross-walk for each FQ (almost no chance → almost certain)
• EEI: analytic confidence (H/M/L) kept separate from likelihood for each FQ
• EEI: date of last probability update and the event / evidence that drove it
• EEI: Brier score or calibration track record for each FQ where a prior version exists

KIQ-2 - Which signposts have fired, are emerging, or are absent?

• EEI: status of each signpost from OSINT-037 §10 (Not present / Emerging / Present)
• EEI: status of each cross-scenario signpost from OSINT-038 §11
• EEI: probability-change magnitude signalled by any fired or emerging signpost
• EEI: lead-time estimate for highest-priority signposts not yet present
• EEI: new potential signposts surfaced during collection not yet in the upstream register

KIQ-3 - Which scenario is the world currently tracking toward?

• EEI: which OSINT-038 scenario the current evidence best matches (scenario discriminators)
• EEI: movement of critical uncertainty axes since last dashboard update
• EEI: qualitative plausibility ordering update across the scenario set
• EEI: any scenario that should be promoted or demoted in plausibility

KIQ-4 - Which horizon-scan watch items have escalated or resolved?

• EEI: current status of each OSINT-039 §8 weak signal (Not present / Emerging / Confirmed / Disconfirmed)
• EEI: any emerging-issues register item (OSINT-039 §6) that has moved priority tier
• EEI: any wildcard / discontinuity (OSINT-039 §10) that has crossed a leading-indicator threshold
• EEI: new weak signals not yet in the upstream horizon scan

KIQ-5 - Have any I&W indicators fired or approached threshold?

• EEI: current state of each OSINT-040 §1 indicator vs. its tripwire threshold
• EEI: L×I severity tier for any indicator at or near tripwire
• EEI: escalation route triggered (if any) and acknowledgement status
• EEI: collection-health check - any OSINT-040 §2 source lane degraded or delayed

PIR-A - Dashboard configuration baseline

• EEI: complete list of active forecast questions (FQs) loaded into the dashboard
• EEI: resolution criteria and resolution date for each FQ
• EEI: complete signpost register cross-linked to FQs and scenario paths
• EEI: upstream product versions currently loaded (OSINT-037, 038, 039, 040 version numbers)

PIR-B - Probability update triggers

• EEI: events or data releases that warrant a dashboard probability re-score
• EEI: cadence-driven refresh schedule (see §00b)
• EEI: escalation threshold - what probability movement triggers client notification

Collection gaps / RFIs (running)

• [gap: FQ probability stale - event occurred after as-of date] → re-score §02 Indicators & Warnings
• [gap: signpost not currently observable] → route to OSINT-040 I&W Program
• [gap: scenario discriminator unclear] → route to OSINT-038 §11 Cross-Scenario Signposts
• [gap: new weak signal requires horizon-scan treatment] → route to OSINT-039 §6–§8
• [gap: base-rate revision warranted] → route to OSINT-037 §7 Reference Class & Base Rates
• [open item]

00b · Monitoring Cadence & Triggers

Continuous dashboard/feed system - this branch drives the live-scoring and alert cadence of the dashboard, not a one-time collection sprint. All upstream products feed this cadence. → governs refresh of §02 Indicators & Warnings, §04 Data Sources & Feeds, §06 Events Timeline.

Active watch items (loaded into dashboard)

• [forecast question - FQ-1 from OSINT-037 §4]
  • Current probability (point + range):
  • ICD 203 band:
  • Resolution date:
  • Last scored:
  • Next scheduled re-score:
• [scenario discriminator signpost - from OSINT-038 §11]
  • Favours scenario:
  • Disfavours scenario:
  • Current status: [Not present / Emerging / Present]
  • Last checked:
• [I&W indicator - from OSINT-040 §1]
  • Tripwire threshold:
  • Current state vs. threshold:
  • Severity tier:
  • Last verified:
• [horizon-scan watch item - from OSINT-039 §8]
  • Signal strength: [Strong / Moderate / Weak]
  • Current status: [Not present / Emerging / Confirmed / Disconfirmed]
  • Last checked:

Refresh cadence

• Routine probability re-score cadence:
• Signpost sweep cadence:
• I&W indicator check cadence:
• Horizon-scan watch-item cadence:
• Full dashboard reconciliation (all upstreams re-synced):
• Calibration/Brier track-record update cadence:

Alert thresholds

• Probability movement threshold for client notification: [e.g., ≥10 percentage-point shift in any FQ]
• Signpost-fired alert threshold: [any signpost moves to Present → immediate notification]
• I&W tripwire alert: [any §1 indicator reaches tripwire → per OSINT-040 §4 severity tier]
• Scenario-tracking alert: [world tracking toward least-plausible scenario → immediate escalation]
• Calibration-degradation alert: [Brier score worsens beyond threshold → re-forecast trigger]

Escalation path

• Probability shift ≥ threshold → [notify named client contact, route to OSINT-037 §16.2 re-forecast]
• Signpost fires (confirming or disconfirming) → [notify analyst, update §02 Indicators & Warnings, re-score FQ]
• I&W tripwire fires → [per OSINT-040 §4 escalation matrix]
• Wildcard/discontinuity leading indicator appears → [escalate to OSINT-038 scenario re-assessment]
• Horizon-scan item reaches High priority → [route to OSINT-039 §16.2 Recommendations]
• Dashboard inputs stale beyond SLA → [flag gap in §99, notify service owner]

01 · Drivers & Key Variables

The structural and proximate forces shaping the forecast questions tracked by the dashboard. Collect current state and trajectory evidence for each driver to power re-scoring decisions. Distinguish slow-moving structural drivers (require infrequent update) from fast-moving proximate ones (require cadence-driven sweep). → feeds OSINT-037 §6 Strategic Drivers & Trends, dashboard re-score logic.

Structural drivers

Long-horizon, slow-moving forces. Update on defined cadence (quarterly / annual data releases). Tools: World Bank Open Data, IMF WEO/IFS, FRED, Eurostat, national statistics portals, ACLED, Uppsala UCDP, V-Dem, Polity5, SIPRI.

• [structural driver - e.g. demographic / economic / institutional / geopolitical trend]
  • Current state (measured):
  • Trajectory (direction + rate): [↑ / → / ↓ + pace]
  • Evidence / data source:
  • Time horizon over which it shapes forecast outcomes:
  • Driver interaction (reinforcing / offsetting / conditional):
  • Forecast relevance: [Dominant / Significant / Marginal]
  • Uncertainty: [High / Med / Low]
  • Confidence: [H/M/L]
  • Source grade (A–F/1–6):
  • Notes / tool output:
• [structural driver 2]

Proximate drivers

Fast-moving, near-term forces. Sweep on routine cadence (daily / weekly). Tools: GDELT GKG, Feedly / RSS aggregation, government press offices, social-media monitoring (lawful / within ToS), prediction market movements (Metaculus, Good Judgment Open).

• [proximate driver - e.g. policy decision, leadership statement, market signal, conflict event]
  • Current state:
  • Trajectory: [↑ / → / ↓]
  • Evidence / event:
  • Interaction with structural drivers:
  • Probability impact on FQ(s): [FQ-1 moves from X% to Y% because…]
  • Forecast relevance: [Dominant / Significant / Marginal]
  • Confidence: [H/M/L]
  • Source grade (A–F/1–6):
  • Notes / tool output:
• [proximate driver 2]

Driver interaction synthesis

• [Reinforcing pair - driver A × driver B]:
• [Offsetting pair - driver A dampens driver B]:
• [Conditional interaction - driver A matters only if driver B exceeds threshold]:

02 · Indicators & Warnings

Observable, monitorable signposts drawn from OSINT-037 §10, OSINT-038 §11, OSINT-039 §8, and OSINT-040 §1 - the live-status layer of the dashboard. Each indicator is a separate clonable block. Collect current status on every refresh sweep. → feeds dashboard Signpost Monitor, OSINT-037 §10, OSINT-040 §1 indicator set. ★ recursive block.

Confirming signposts (would raise central-forecast probability)

Source: OSINT-037 §10 - confirming signposts. Clone per signpost.

• [signpost - e.g. observable event, metric threshold, policy action, official statement]
  • Source product & section: [OSINT-037 §10 / OSINT-038 §11]
  • Forecast question it bears on: [FQ-1 / FQ-2 / FQ-3]
  • Probability change if observed: [e.g. FQ-1: 55%→75%]
  • Lead time:
  • Current status: [Not present / Emerging / Present]
  • Last checked (date):
  • Where to monitor (source / URL):
  • Disposition: [Watch / Notify client / Re-score / Route to OSINT-040 I&W]
  • Notes / tool output:
• [signpost 2]

Disconfirming signposts (would lower central-forecast probability)

Source: OSINT-037 §10 - disconfirming signposts. Clone per signpost.

• [signpost - e.g. non-event, policy reversal, contradiction of driver trajectory]
  • Source product & section: [OSINT-037 §10 / OSINT-038 §11]
  • Forecast question it bears on: [FQ-1 / FQ-2 / FQ-3]
  • Probability change if observed: [e.g. FQ-1: 55%→35%]
  • Lead time:
  • Current status: [Not present / Emerging / Present]
  • Last checked (date):
  • Where to monitor:
  • Disposition: [Watch / Route to OSINT-040 I&W / Re-score]
  • Notes / tool output:
• [signpost 2]

I&W tripwire indicators (from OSINT-040 §1)

Each I&W indicator loaded from the standing OSINT-040 program. Clone per indicator.

• [indicator - from OSINT-040 §1 indicator register]
  • Observable dimension:
  • Tripwire threshold:
  • Current state vs. threshold:
  • Likelihood (1–5):
  • Impact (1–5):
  • L×I score (1–25):
  • Severity tier: [Critical / High / Elevated / Moderate / Low]
  • Current status: [Within bounds / Approaching threshold / Threshold reached / Fired]
  • Last verified (date):
  • Escalation route (→ OSINT-040 §4):
  • Notes / tool output:
• [indicator 2]

Wildcard / discontinuity tripwires

Low-probability, high-impact signals from OSINT-038 §9 and OSINT-039 §10. Clone per item.

• [wildcard - e.g. exogenous shock leading indicator, black-swan-adjacent signal]
  • Source product & section: [OSINT-038 §9 / OSINT-039 §10]
  • Type: [Discontinuity / Wildcard / Black-swan-adjacent]
  • Estimated probability: [e.g. <5%]
  • Impact if triggered: [Critical / High]
  • Leading indicators to watch:
  • Current status: [Absent / Trace signal / Strengthening]
  • Last checked:
  • Disposition: [Watch / Escalate to OSINT-038 scenario re-assessment]
  • Notes:

03 · Scenarios / Hypotheses

The calibrated forecast outcomes and scenario paths the dashboard tracks and scores. Collect evidence bearing on each hypothesis for ACH and live re-scoring. → feeds dashboard Forecast Question Register (live tracking), OSINT-037 §8 The Forecast, OSINT-038 §7–§9 Scenario Register, OSINT-037 §12 ACH, OSINT-037 §14 Premortem.

Active forecast questions (FQ register - live)

One block per active FQ from OSINT-037 §4 / §8. Clone per question.

• [FQ-1 - resolvable question from OSINT-037 §4, e.g. “Will [outcome] occur by [date]?“]
  • Resolution criteria:
  • Resolution date:
  • Current point probability:
  • Current range:
  • ICD 203 band:
  • Analytic confidence: [H/M/L]
  • Base-rate anchor (outside view):
  • Case-specific adjustment (inside view):
  • Last scored (date):
  • Evidence FOR at current collection date (C):
    • [evidence item]:
  • Evidence AGAINST at current collection date (I):
    • [evidence item]:
  • Neutral evidence (N):
    • [evidence item]:
  • Signposts fired / emerging since last score:
  • Notes / Brier running score:
• [FQ-2]
• [FQ-3]

Scenario tracker (from OSINT-038 §7–§9)

One block per scenario. Clone per scenario. Update which scenario the current evidence tracks toward.

• [Scenario name - from OSINT-038 §7 Scenario Register]
  • Defining logic (axis combination):
  • Current plausibility assessment: [Most plausible / More likely than not / Roughly even / Less likely / Least plausible]
  • Previous plausibility:
  • Direction of movement: [Strengthening / Stable / Weakening]
  • Evidence driving the movement:
  • Discriminating signposts now present:
  • Discriminating signposts absent (should be present if tracking):
  • Client implications if tracking:
  • Notes:
• [Scenario 2]

Key Assumptions Check (KAC) - live refresh

Assumptions loaded from OSINT-037 §13. Check each on cadence for changed basis. → §07 Assumptions & Uncertainties.

• [assumption - e.g. reference class is genuinely comparable]
  • Basis / evidence (current):
  • Confidence: [H/M/L]
  • Any change since last review?: [Y/N - describe if Y]
  • Impact on forecast if wrong:
  • Linchpin?: [Y/N]
  • Notes:
• [assumption - dominant driver continues current trajectory]
  • Basis / evidence (current):
  • Any change?: [Y/N]
  • Impact if wrong:
  • Linchpin?: [Y/N]

Premortem / failure-mode register (live)

Failure modes from OSINT-037 §14. Review on each major refresh. → §07 Assumptions & Uncertainties.

• [failure mode - what would cause the central forecast to be badly wrong]
  • Assumption / driver that would break:
  • Likely bias: [Anchoring / Confirmation / Base-rate neglect / Recency / Optimism]
  • Current early-catch status: [Check installed / Not installed - see §02 Indicators]
  • Notes:

04 · Data Sources & Feeds

Every source and feed powering the dashboard - institutional, statistical, event data, prediction market aggregators, and human sourcing. Cadence-graded and reliability-graded. This is the dashboard’s living data-ingest layer; sources should be reviewed on each scheduled refresh and re-graded when access or quality changes. → feeds dashboard configuration, OSINT-037 §17 Annex A, Appendix H Full Source Register. ★ recursive block.

Institutional / official statistical feeds

Tools: World Bank Open Data, IMF WEO/IFS/Article IV, UN Statistics, Eurostat, FRED, OECD.Stat, central bank portals, national statistics agency portals.

• [feed - e.g. IMF World Economic Outlook / World Bank Databank / national statistics]
  • Metric / series tracked:
  • FQ(s) / driver(s) it informs:
  • Cadence / release schedule:
  • Self-interest / reliability caveat:
  • Reliability (A–F):
  • Credibility (1–6):
  • Access URL:
  • Notes / tool output:
• [feed 2]

Conflict / political event datasets

For base-rate reference-class collection and event-timeline maintenance. Tools: ACLED, Uppsala UCDP, GDELT, ICEWS, V-Dem, Polity5, PRS Group, IISS Military Balance, Jane’s / Janes.com, SIPRI, ICG Crisis Watch.

• [dataset - e.g. ACLED / UCDP / V-Dem / GDELT GKG]
  • Events / metrics tracked:
  • FQ(s) / driver(s) it informs:
  • Cadence / data currency:
  • Reference class supported:
  • Reliability (A–F) / Credibility (1–6):
  • Access route / licence:
  • Notes:
• [dataset 2]

Prediction market & forecasting aggregator feeds

Community probabilities as calibration cross-check - not ground truth. Compare to internal FQ scores. Tools: Metaculus (metaculus.com), Good Judgment Open (gjopen.com), Manifold Markets (manifold.markets), Polymarket (on-chain prices as market signal only), RAND Pardee forecasting.

• [platform - e.g. Metaculus / Good Judgment Open / Manifold]
  • Question URL / title:
  • FQ mirrored (internal): [FQ-1 / FQ-2 / FQ-3]
  • Current community probability:
  • Resolution criteria on platform:
  • Last checked (date):
  • Deviation from internal forecast (±):
  • Disposition if significant deviation: [Flag for review / Investigate driver]
  • Notes:
• [platform 2]

Open-source news & media monitoring feeds

Sweep on routine cadence for signpost and driver-state updates. Tools: GDELT GKG, Feedly, Google News RSS, Nexis/Factiva (licenced), OSINT operators (site:, filetype:, date:, inurl:), Media Cloud, RavenDB open-source media indices.

• [feed / outlet - e.g. Reuters / AP / regional outlet / aggregator / GDELT GKG]
  • Coverage:
  • FQ(s) / signposts monitored:
  • Cadence:
  • Self-interest / state-control caveat:
  • Reliability (A–F) / Credibility (1–6):
  • Notes:
• [feed 2]

Academic / think-tank / analytical feeds

Lower cadence; sweep weekly or on publication. Tools: Google Scholar alerts, SSRN new papers, Brookings, RAND, Chatham House, IISS, ICG, CSIS, SWP, ECFR, CFR, regional area-studies journals.

• [publication / institution]
  • Relevance to FQ(s) / driver(s):
  • Alert / cadence:
  • Reliability (A–F) / Credibility (1–6):
  • Notes:

Human / in-country sourcing (where commissioned)

Expert network, analyst contacts, Track-II sourcing, in-country correspondents.

• [source type / designator - never name if sensitive]
  • Reliability (A–F):
  • Credibility (1–6):
  • FQ(s) / driver(s) informed:
  • Last contact:
  • Notes:

05 · Key Actors & Entities

The actors whose decisions or behavior most move forecast probabilities tracked by the dashboard. Clone per actor. Monitor actor posture and statements on each cadence sweep. → feeds dashboard re-score logic, OSINT-037 §6 Strategic Drivers (actor as proximate driver), OSINT-037 §10 Signposts (actor behavior as signpost), OSINT-038 §8 Scenario Narratives. ★ recursive block.

State / government actors

Tools: government press offices, official statement portals, Hansard / congressional records, diplomatic statements, UN General Debate speeches, FOIA releases; OpenCorporates for state-owned enterprises; sanctions registers (OFAC SDN, EU Consolidated, UN Consolidated).

• [actor - e.g. government, ministry, ruling party, central bank, military command]
  • Role in dashboard forecasts:
  • FQ(s) most affected by this actor’s decisions: [FQ-1 / FQ-2 / FQ-3]
  • Current posture / stated position:
  • Trajectory / trend in behavior: [↑ / → / ↓]
  • Decision levers relevant to FQs:
  • Constraints on action:
  • Confirming signpost: what actor behavior would raise FQ probability:
  • Disconfirming signpost: what actor behavior would lower FQ probability:
  • Last monitored (date):
  • Source / basis:
  • Confidence: [H/M/L]
  • Notes / tool output:
• [actor 2]

Non-state actors

Opposition groups, armed groups, civil society, major corporations, multilateral bodies, IFIs. Tools: UN Panel of Experts reports, NGO databases (ACLED actor list, Crisis Group actor profiles), OFAC/EU/UN sanctions registries, OpenCorporates, company registries.

• [actor - e.g. opposition party, armed faction, IGO, major firm, lobby group]
  • Role in dashboard forecasts:
  • FQ(s) affected: [FQ-1 / FQ-2]
  • Current posture:
  • Trajectory: [↑ / → / ↓]
  • Constraints:
  • Confirming signpost:
  • Disconfirming signpost:
  • Last monitored (date):
  • Source / basis:
  • Confidence: [H/M/L]
  • Notes / tool output:
• [actor 2]

External / third-party actors

Foreign states or transnational actors whose intervention could move forecast probabilities.

• [actor - e.g. foreign power, multilateral institution, diaspora network]
  • Stake in dashboard forecast outcomes:
  • Likely posture / intervention vector:
  • Evidence of current activity:
  • Probability impact on FQs if they act:
  • Last monitored (date):
  • Source:
  • Confidence: [H/M/L]
  • Notes:

06 · Events Timeline

Chronological record of events bearing on drivers, signposts, or forecast questions - the live event log that powers dashboard re-scoring decisions. Clone per event. On each cadence sweep, add new events to the current-period sub-block. → feeds dashboard Forecast Question Register (re-score triggers), OSINT-037 §11 Key Findings, Appendix G Time-Phasing Chart, OSINT-037 §6 driver trajectories. ★ recursive block.

Past events (pre-collection / pre-dashboard-launch)

Events establishing baseline, reference-class cases, and driver trajectory.

• [event - date + brief description]
  • Date:
  • Description:
  • Bearing on: [driver / signpost / reference class / alternative outcome / scenario path]
  • Forecast question(s) affected: [FQ-1 / FQ-2 / FQ-3]
  • Probability impact (retrospective):
  • Source:
  • Source grade (A–F/1–6):
  • Notes:
• [event 2]

Current-period events (within dashboard monitoring window)

Events collected on routine cadence that may trigger a re-score or alert.

• [event - date + brief description]
  • Date:
  • Description:
  • Bearing on: [driver / signpost / forecast probability / scenario discriminator]
  • Probability impact: [FQ-1 moves from X% to Y% because…]
  • Re-score triggered?: [Y/N]
  • Alert triggered?: [Y/N - tier if yes]
  • Source:
  • Source grade (A–F/1–6):
  • Notes:
• [event 2]

Scheduled / anticipated future events (within forecast horizon)

Elections, summits, contract renewals, fiscal deadlines, leadership transitions - potential signpost resolution dates or probability-shift triggers.

• [anticipated event - expected date + type]
  • Expected date:
  • Type: [Election / Summit / Deadline / Leadership transition / Economic release / Legal deadline]
  • Forecast question(s) affected: [FQ-1 / FQ-2]
  • Signpost implication: [If X occurs → probability moves to Y]
  • Scenario implication: [Confirms / Disfavours scenario name]
  • Source / basis for timing:
  • Monitor via:
  • Notes:
• [anticipated event 2]

Calibration track record (resolution events)

Record of FQs that have resolved - for Brier scoring and calibration audit. → feeds Appendix F Calibration & Track Record in OSINT-037.

• [resolved FQ - question text]
  • Resolution date:
  • Resolution criteria met?: [Y/N - describe]
  • Internal forecast at resolution:
  • Actual outcome:
  • Brier score contribution:
  • Calibration lesson:
  • Notes:

07 · Assumptions & Uncertainties

The analytic assumptions and irreducible uncertainties underpinning all active forecasts on the dashboard. Loaded from OSINT-037 §13 KAC and refreshed on each scheduled review. Also captures the bias audit run before each re-score. → feeds OSINT-037 §5 Forecast Framing & Methodology, §13 KAC, §14 Premortem, dashboard Calibration/Brier Track Record configuration.

Reference-class and base-rate assumptions

• [assumption - e.g. reference class for FQ-1 is genuinely comparable]
  • Basis / evidence (current):
  • Confidence: [H/M/L]
  • Changed since last review?: [Y/N]
  • Impact on FQ probability if wrong:
  • Linchpin?: [Y/N]
  • Notes:

Driver-continuity assumptions

• [assumption - dominant driver continues trajectory through horizon]
  • Basis / evidence (current):
  • Confidence: [H/M/L]
  • Changed since last review?: [Y/N]
  • Impact if wrong:
  • Linchpin?: [Y/N]
  • Notes:

Exogenous-shock and data-quality limitations

• [limitation - e.g. base-rate sample small; no major exogenous shock assumed]
  • Nature of limitation:
  • Effect on FQ range (widening required?):
  • Mitigation:
  • Notes:

Reflexivity and model uncertainty

• [reflexivity concern - e.g. dashboard publication or client action could alter outcome]
  • Nature:
  • Assessed probability of material distortion:
  • Mitigation:

Bias audit (run before each re-score)

Structured check before committing any updated probability to the dashboard.

• Anchoring: [Were probabilities adjusted sufficiently from the anchor?]:
• Confirmation bias: [Was disconfirming evidence genuinely weighed?]:
• Base-rate neglect: [Was the outside view applied before the inside view?]:
• Recency bias: [Were vivid recent events over-weighted vs. base rate?]:
• Optimism / pessimism: [Was the range widened appropriately for residual uncertainty?]:

Dashboard-specific uncertainty register

Uncertainties not in OSINT-037 §13 that are specific to the dashboard’s live-tracking role.

• [uncertainty - e.g. source-feed latency; prediction-market thin liquidity distorting signal]
  • Nature:
  • FQ(s) affected:
  • Monitoring response:
  • Notes:

99 · Collection Admin

Working register - audit trail behind the dashboard, not a deliverable section.

Source register

Every material datum traceable to a graded source. Admiralty two-axis code (A–F / 1–6).

• [S-1 - source name / designator]
  • Type: [Primary official / Secondary / Tertiary / Dataset / Prediction market / HUMINT]
  • Reliability (A–F):
  • Credibility (1–6):
  • Date accessed:
  • Coverage scope:
  • Self-interest / caveat:
• [S-2]
• [S-3]

Evidence archive

• [capture ref + hash + URL + timestamp + description]:

Forecast question register (working - live scores)

Master working list of all active FQs loaded into the dashboard.

• [FQ-1 - resolvable question text]
  • Resolution criteria:
  • Resolution date:
  • Current probability (working):
  • Range:
  • ICD 203 band:
  • Analytic confidence: [H/M/L]
  • Last scored (date):
  • Last event driving a re-score:
• [FQ-2]
• [FQ-3]

Upstream product version log

Track which version of each upstream product the dashboard currently reflects.

• OSINT-037 Strategic Forecast & Estimate: [version + date loaded]
• OSINT-038 Scenario / Alternative Futures Analysis: [version + date loaded]
• OSINT-039 Horizon Scan / Emerging Threats Report: [version + date loaded]
• OSINT-040 I&W Program: [version + date loaded / active term]

Open gaps / RFIs (running)

• [base-rate source gap - pending] → OSINT-037 §7 Reference Class & Base Rates / OSINT-029 Intelligence Estimate
• [signpost not observable - standing watch needed] → OSINT-040 I&W Program
• [driver state contested - corroboration needed] → OSINT-026 Country / Regional Study
• [full alternative-futures update warranted] → OSINT-038 Scenario / Alternative Futures Analysis
• [new weak signal requiring horizon-scan sweep] → OSINT-039 Horizon Scan / Emerging Threats Report
• [FQ resolution criteria ambiguous - clarification needed] → client / OSINT-037 analyst
• [calibration degradation - re-forecast warranted] → OSINT-037 §16.2 Recommendations
• [item submitted, not yet returned]