1) Diagnostic Identity

Diagnostic Name: Attention Capacity

Short Name / Symbol: attention_capacity

Diagnostic Class: Attention / Observation / Processing Capacity / Ψ Stability / Cognitive Throughput

Primary Function: Estimate how much meaningful signal, complexity, responsibility, feedback, memory, risk, context, or environmental change a system can attend to without losing coherence, misclassifying reality, collapsing decision depth, or allowing important signal to disappear.

Primary Use: Determine whether the system has enough attention available to observe, interpret, prioritize, repair, and remember what matters.

Core Risk if Ignored: The system may receive signal but fail to notice, hold, process, or respond to it, producing misclassification, delayed response, hidden debt, false closure, and attention-driven coherence loss.

Core Risk if Overtrusted: Attention may be mistaken for infinite awareness, causing the system to take on more signals, responsibilities, or risks than it can meaningfully process.

2) Mechanical Definition

attention_capacity measures the usable amount of focused, coherent attention available for reality-contact, signal interpretation, repair, memory, and decision-making.

attention_capacity answers:

Can the system actually attend to what it is responsible for seeing?

Attention is not merely awareness.

A system can technically receive information while lacking capacity to attend to it.

Attention capacity includes the ability to:

notice signal
hold context
compare alternatives
track recurrence
preserve boundary signals
detect weak signals
distinguish noise from meaning
maintain decision depth
remember prior evidence
route feedback
notice affected-node cost
notice hidden debt

Attention capacity is one of the practical foundations beneath Ψ Presence / Attention.

A simple form:

attention_capacity = usable observation + context-holding + prioritization + recall bandwidth

When attention capacity falls below signal load:

signal_load > attention_capacity ⇒ missed-signal debt ↑

3) What the Diagnostic Measures

Direct Measurement Target

attention_capacity measures:

• observation capacity
• signal-holding capacity
• context retention capacity
• prioritization capacity
• complexity attention capacity
• feedback attention capacity
• affected-node signal visibility
• weak-signal detection
• recurrence tracking capacity
• memory recall during decision
• attention available for repair
• attention available for boundary signals
• attention available for contradiction
• attention available for slow variables
• attention available for cross-layer interpretation
• whether the system can notice what it claims to govern

Indirect / Proxy Signals

attention_capacity can be estimated from:

• missed signals
• delayed recognition
• repeated “we did not see it”
• shallow summaries replacing analysis
• weak signals being ignored
• affected-node reports getting lost
• recurring issues treated as new
• high context switching
• decision compression
• backlog of unread feedback
• overreliance on dashboards
• high urgency reducing observation
• important details disappearing in handoff
• inability to track multiple variables in S
• attention captured by Φ while H rises
• high noise causing meaningful signal loss
• audit trails existing but not being read
• memory present but not retrieved at decision time

What It Does Not Measure

attention_capacity does not directly measure:

• intelligence
• intent
• care
• truthfulness
• morality
• full system competence
• whether the signal itself is valid
• whether the system has enough repair capacity
• whether every signal should receive equal attention
• whether attention should be unlimited
• whether all ignored signal was important
• whether high attention always produces correct action

High attention_capacity means the system can observe and process more meaningful reality before compression or omission occurs.

It does not guarantee correct interpretation.

Low attention_capacity means important signal may be missed, flattened, delayed, or misrouted.

It does not mean the system is unwilling to attend.

4) Canonical State Variables Involved

Canonical state vector:

S = {O, H, ε, ι, Au, µᵢ, BΣ, K, R, Φ}

Primary Variables

• O: coherence depends on sufficient attention to real conditions
• H: hidden debt rises when signal is missed or not held long enough
• Au: auditability is only useful if the system can attend to the trace
• R: restoration depends on noticing what must be repaired
• BΣ: boundary strain often requires attention before breach
• µᵢ: integrity depends on perceiving effects of action accurately

Secondary Variables

• ε: visible error competes for attention and can crowd out hidden debt
• ι: pseudo-coherence rises when attention is captured by surface order
• K: compatibility requires attention to both nodes, not only the dominant signal
• Φ: proxy metrics can capture attention away from O and H

Variables Commonly Confused With attention_capacity

5) Localization Signature

Primary Legibility Layers

• U1 — Power / Budgets: attention is a limited budget of time, energy, staffing, compute, and cognitive bandwidth
• U3 — Execution: where attention is spent on tasks, monitoring, response, and practical operations
• U4 — Classification / Metrics / Narratives: where attention is directed by categories, dashboards, priorities, and stories
• U5 — Coordination / Time: where attention is scheduled, sequenced, interrupted, delayed, or overloaded
• U6 — Coherence Field: where shared attention determines what the system collectively notices
• U7 — Memory / Recurrence: where attention must retrieve past patterns and warnings
• U8 — Environment / Forcing: where attention is pulled by external stress, novelty, crisis, or noise

Primary Leverage Layers

• U1: increase attention resources, slack, staffing, compute, or time
• U3: reduce task overload and create observation routines
• U4: repair attention-directing categories and metrics
• U5: protect time windows for review, reflection, and recurrence tracking
• U6: restore shared attention to reality, not only urgency
• U7: improve retrieval of past signal during decisions

Verification Layers

• U1: does the system have enough attention budget?
• U3: are important signals being observed in practice?
• U4: are metrics/narratives directing attention correctly?
• U5: is there time to attend before response is required?
• U6: does the collective field notice the right things?
• U7: are past signals remembered at decision time?

Common Mislocalizations

• Treating signal absence as issue absence
• Treating unread feedback as low feedback
• Treating dashboard attention as reality attention
• Treating crisis attention as repair attention
• Treating awareness as processing
• Treating processing as response
• Treating attention failure as lack of care
• Treating high urgency as high importance
• Treating loud signals as highest priority
• Treating low-rank signal as low relevance
• Treating memory existence as memory use
• Treating formal monitoring as actual attention

6) Input Requirements

Required Inputs

To estimate attention_capacity, the system needs:

• signal field being evaluated
• current signal load
• attention resources available
• affected variables in S
• current urgency load
• feedback backlog
• monitoring pathways
• memory retrieval capacity
• prioritization criteria
• attention allocation by domain
• affected-node signal visibility
• weak-signal pathways
• decision cadence
• current Φ pressure
• current H indicators
• response or review backlog

Optional Inputs

These improve precision:

• task load
• context-switching rate
• staffing / compute / time budget
• meeting/review cadence
• alert volume
• noise level
• signal-to-noise ratio
• feedback queue age
• unread report count
• escalation delays
• missed-warning history
• recurrence history
• audit-log access records
• attention distribution map
• dashboard usage
• affected-node reporting patterns
• crisis load
• memory retrieval logs
• decision-quality postmortems

Missing Input Behavior

If attention_capacity inputs are missing:

• If signal load is unknown, attention sufficiency cannot be judged
• If attention allocation is unknown, assume important domains may be undersampled
• If feedback backlog is unknown, FI may be overestimated
• If memory retrieval is unknown, recurrence may be missed
• If affected-node visibility is unknown, cost-bearing nodes may be unattended
• If Φ pressure is high, check whether attention is captured by metrics
• If urgency load is high, assume decision depth may be compressing
• If H indicators are unknown, visible attention may be missing slow debt

Default missing-input posture:

map signal load → map attention allocation → compare to risk/consequence → protect review and memory pathways

7) Diagnostic States / Ranges

These ranges are qualitative and should be domain-calibrated.

Healthy / Coherence-Supporting Range

The system can attend to meaningful signal without excessive compression, omission, or misclassification.

Signals:

• signal load is manageable
• feedback is reviewed in time
• weak signals have pathways
• affected-node reports are seen
• memory is retrieved during decisions
• prioritization is explicit
• attention is not fully captured by crisis or Φ
• boundary strain is noticed early
• recurrence patterns are recognized
• attention supports repair and adaptation

Recommended posture:

continue observation routines
preserve attention slack
monitor signal load
update U7 recurrence memory

Watch Range

Attention load is rising and important signals may begin slipping.

Signals:

• feedback review slows
• weak signals receive less attention
• summaries replace source too quickly
• context windows narrow
• affected-node reports wait longer
• recurrence recognition is inconsistent
• urgent tasks crowd out review
• dashboards dominate attention
• attention is available only for visible ε
• H indicators are under-reviewed

Recommended posture:

reduce noise
prioritize high-consequence signal
increase review capacity
protect attention for H/BΣ/FI

Degraded Range

Attention capacity is insufficient for the system’s signal load and consequence level.

Signals:

• repeated missed warnings
• feedback backlog grows
• affected-node signal is lost
• recurrence is treated as new
• hidden debt rises unnoticed
• boundary strain is noticed after breach
• decision depth collapses under load
• attention is captured by metrics or crisis
• memory is not retrieved
• low-rank or quiet signal disappears
• repair starts late because signal was missed

Recommended posture:

pause expansion
reduce signal load
increase attention resources
triage signal by consequence
restore memory and feedback review

Contraindicated:

scaling responsibility
high-impact actuation
declaring no issue from no visible signal
adding metrics without attention review
deep coupling with unattended domains

Critical / Collapse-Prone Range

The system cannot attend to essential reality-contact and is operating blind or in forced-response mode.

Signals:

• major signals are missed until crisis
• hidden debt becomes active failure
• feedback systems are unread or ignored
• boundary breaches recur unnoticed
• memory is functionally unavailable
• crisis attention consumes all capacity
• system cannot distinguish signal from noise
• affected nodes exit or stop reporting
• decisions are made from compressed fragments
• external audit is needed to reconstruct what was missed

Recommended posture:

stop nonessential commitments
restore minimal attention capacity
protect critical feedback channels
reduce noise and urgency
rebuild monitoring and U7 retrieval
triage by risk and affected-node cost

False Positive Risk

attention_capacity may appear low when:

• the system is intentionally filtering noise
• low-priority signals are correctly deferred
• attention is concentrated during a legitimate crisis
• slow review reflects careful processing
• signals are being batched efficiently
• attention has shifted to origin-layer repair
• visible reduction in attention is actually reduced signal load
• automation is correctly handling low-risk signals

False Negative Risk

attention_capacity may appear high when:

• dashboards look complete but miss affected-node reality
• many signals are received but not processed
• feedback is acknowledged but not understood
• reports are skimmed or summarized poorly
• high-status signal crowds out low-status signal
• attention is captured by Φ
• crisis response creates illusion of vigilance
• memory exists but is not retrieved
• quiet nodes have stopped reporting

8) Leading Indicators

attention_capacity degradation appears early as:

• review windows shrink
• feedback waits longer
• people summarize without reading source
• “we missed that” repeats
• weak signals disappear
• affected-node reports become stale
• recurrence is rediscovered
• dashboards become primary reality
• urgent issues crowd out important issues
• decision notes lose nuance
• attention shifts to only visible ε
• memory references become vague
• context has to be reloaded repeatedly
• noise increases faster than filtering
• boundary strain is only noticed late

9) Lagging Indicators

attention_capacity failure has already accumulated debt when:

• crisis reveals ignored warnings
• external audit finds missed evidence
• affected nodes disengage
• hidden debt surfaces suddenly
• repeated issues were documented but unread
• repair starts after avoidable damage
• official memory is incomplete
• legitimacy shock follows “we should have known”
• system cannot reconstruct missed signal
• attention collapse becomes normal
• high-consequence decisions were made from shallow context
• slow variables become active failure

10) Interpretation Rules

How to Read attention_capacity

attention_capacity should be read as:

usable coherent attention relative to signal load and consequence severity

It is not raw awareness or good intent.

A system may have:

• high attention capacity and low signal load
• high attention capacity but poor prioritization
• low attention capacity but stable operation under low complexity
• high signal intake and low actual attention
• high crisis attention and low repair attention
• strong metric attention and weak affected-node attention
• strong U3 attention and weak U7 memory attention

What Changes Its Meaning

attention_capacity changes meaning under:

• high signal load
• high consequence severity
• high Cv(t)
• high AP(t)
• high Φ pressure
• high X_c(t)
• high crisis_loop_index
• high stress_divergence
• weak FI_integrity
• low EB
• low Au_eff
• low M_int(t)
• short τ_m(t)
• high boundary_strain
• high affected_node_cost
• high U8 forcing

Context Modifiers

High signal load: attention must increase or prioritization must tighten.

High consequence severity: missed signal becomes more costly.

High Cv(t): rapid compression reduces attention depth.

High Φ pressure: metrics may capture attention away from O.

High X_c(t): rule complexity consumes attention.

Weak FI: feedback may not receive enough attention to correct.

Low EB: low signal may reflect expression limits, not low need.

Low M_int(t): memory cannot support attention across time.

High U8 forcing: external stress can hijack attention.

Domain Calibration Notes

attention_capacity should be calibrated by domain:

• in engineering: alert load, incident monitoring, code review depth, postmortem attention, dependency tracking
• in AI: context window use, retrieval attention, tool-result inspection, user feedback review, safety signal triage
• in institutions: complaint review, case load, staff attention, audit review, affected-node tracking
• in governance: public signal processing, oversight capacity, crisis attention, long-term policy attention
• in relationships: ability to attend to boundary signals, repair memory, timing, and repeated patterns
• in archives: ability to track glossary drift, cross-links, canon status, source lineage, and reader confusion

11) Operator Sequencing Implications

If attention_capacity Is Healthy

Allowed with ordinary gate checks:

• Ψ attention can support Μ sensemaking
• FI feedback can be processed
• Γ selection can use broader signal field
• ℛ repair can be targeted earlier
• U7 memory can be retrieved during decisions
• Δ tests can be interpreted reliably
• Τ trajectory can proceed with monitoring

Recommended:

Ψ attend → Μ interpret → Γ prioritize → ℛ repair → U7 memory update

If attention_capacity Is Low

Recommended:

pause expansion → reduce signal/noise load → triage by consequence → restore review and memory capacity → then decide

Or:

protect critical attention channels for H, BΣ, affected-node cost, FI, and recurrence

Avoid or delay:

• high-impact actuation
• irreversible Π
• deep coupling
• scaling responsibility
• declaring no issue from no visible signal
• rapid Τ acceleration
• metric-only decision-making
• durable memory binding from shallow review

Operators Recommended Under Low attention_capacity

• Ψ: restore direct attention
• Θ: damp urgency and certainty
• Γ: prioritize signal fields
• Π: reduce attention load and constrain noise
• Au: make key traces easier to inspect
• FI: protect feedback pathways
• ℛ: repair attention infrastructure
• Μ: rebuild context before selection

Operators Contraindicated Under Low attention_capacity

• Γ hard selection: may select from incomplete signal
• Π irreversible constraint: may encode missed context
• ⊗ deep coupling: increases signal and dependency load
• ⊕ composition: embeds unprocessed complexity
• Τ acceleration: outruns attention
• Σ escalation: sacralizes shallow reading
• ✕ force: often suppresses signal and increases hidden debt

12) Gate Implications

Gates Strengthened By Reliable attention_capacity

• FI-Gate: feedback can actually be received and reviewed
• Au-Actuation: audit traces are usable because they are attended to
• High Risk Gate: blocks binding when attention is too compressed
• MS-Gate: checks whose signals are attended to or ignored
• ☷ᵢ: ensures principles are applied with sufficient context

Gates Weakened If attention_capacity Is Poor or Unknown

If attention capacity is low:

• FI may collect feedback that no one processes
• Au may exist without use
• High Risk Gate may bind classifications from shallow review
• MS may miss low-visibility affected nodes
• ☷ᵢ may become sloganized due to low context
• Π may constrain from incomplete signal
• Γ may select loud or metric-friendly options
• ℛ may repair late or at the wrong layer

Gate Outcomes Affected

Low attention_capacity should push gates toward:

• Pause
• Reduce signal load
• Require review capacity
• Require affected-node signal check
• Require memory retrieval
• Require source inspection
• Deny high-risk binding
• Deny metric-only closure
• ∅ for high-impact action when the system cannot attend to relevant signal fields

13) Scaling Behavior

attention_capacity becomes harder under scale because signal volume, complexity, noise, feedback, memory, and coordination all increase.

As systems scale:

• signal volume rises
• noise rises
• dashboards multiply
• feedback queues grow
• weak signals disappear
• affected-node reports are summarized away
• attention becomes role-fragmented
• slow variables lose attention
• memory retrieval becomes harder
• crisis attention dominates
• high-status signal crowds out low-status signal
• proxy metrics direct attention
• decision depth compresses
• attention becomes the scarce governance resource

Scaling Risks

• attention capture
• missed-signal debt
• hidden debt accumulation
• boundary signal loss
• feedback theater
• dashboard blindness
• shallow decision-making
• recurrence misrecognition
• affected-node invisibility
• crisis-driven attention
• long-term neglect
• legitimacy shock from ignored warnings
• memory non-use
• signal-to-noise collapse
• forced-response governance

Scaling Requirements

To scale attention safely, systems need:

• attention budgets
• signal triage
• noise filtering
• affected-node signal pathways
• weak-signal channels
• review cadence
• dashboard scope notes
• source inspection routines
• memory retrieval systems
• recurrence tracking
• slow-variable monitoring
• alert hygiene
• feedback queue limits
• consequence-based prioritization
• attention audits
• attention redundancy for high-risk domains

Scaling Rule

Attention capacity must scale with signal load, consequence severity, complexity, and hidden-debt risk.

Sanity constraint:

signal_load > attention_capacity ⇒ missed_signal_debt ↑

If incoming signal exceeds attention, important reality-contact is lost.

Second constraint:

attention_capacity ↓ + High Risk Gate binding ↑ ⇒ downstream error risk ↑

If high-risk binding occurs under low attention, improper binding risk rises.

Third constraint:

Φ_attention ↑ + O/H_attention ↓ ⇒ Goodhart blindness risk ↑

If attention follows proxy metrics while coherence and hidden debt are unattended, Goodhart blindness rises.

14) Interaction / Coupling Behavior

attention_capacity reveals whether a relation, institution, AI system, archive, or interface can actually attend to the reality of all coupled nodes.

What It Reveals About Coupling

• whether one node’s signal is consistently missed
• whether loud signals dominate quiet signals
• whether feedback is heard or merely received
• whether boundary strain is noticed before rupture
• whether repair requires repeated reminders
• whether one node must manage the other’s attention
• whether coupling creates more signal than either node can process
• whether shared attention can hold complexity

What It Reveals About Boundary Integrity

Boundary signals often require attention before they become visible breaches.

When attention capacity is low:

• refusal may be missed
• consent ambiguity may persist
• boundary strain may be recognized late
• repeated clarification becomes necessary
• BΣ repair begins only after rupture
• affected-node cost is undercounted
• quiet boundaries become overwritten by louder signals

What It Reveals About Compatibility

Compatibility requires mutual attention capacity.

A coupling may be unsafe if:

one node must repeat signal many times before the other notices

or:

the coupling creates more complexity than the shared attention field can hold

Healthy compatibility includes enough attention to notice, remember, and respond to each node’s reality.

Relevant Interface Acts

• Ψ Presence / Attention: primary operator support
• ↺ Reflection: confirms that signal was actually received
• ⇩ Relaxation: reduces urgency and attention compression
• ⊘ Attenuation: reduces coupling load when attention is insufficient
• ⊙ Alignment: checks whether one is attending to one’s own role and effects
• →? Invitation: invites signal without overloading the channel
• ⚕︎ Restorative Override: requires post-action attention review
• ✕ Force: often suppresses signal and overloads attention

15) Failure Modes Detected

Primary Failure Modes

attention_capacity detects or predicts:

• missed-signal debt
• feedback backlog
• boundary signal loss
• weak-signal loss
• affected-node invisibility
• dashboard blindness
• shallow decision-making
• memory non-use
• recurrence misrecognition
• crisis attention capture
• proxy attention capture
• delayed repair
• context collapse
• classification error from shallow review
• attention exhaustion
• slow-variable neglect
• forced-response governance

Composite Regimes Where attention_capacity Matters

• Compression Collapse: attention narrows and decision depth falls
• Goodhart Collapse: attention captured by Φ
• Crisis Loop: crisis consumes attention and prevents repair
• Repair Theater: visible repair gets attention while hidden debt does not
• Pseudo-Coherent Basin: system attends to order signs and misses H
• Mission Lock: attention narrows around trajectory
• Taboo Lock: attention avoids protected zones
• Extraction Regime: cost-bearing nodes receive less attention
• LOS: actual operation is unattended because formal map captures attention

16) Accountability & Reintegration Implications

If attention_capacity Was Ignored

Likely consequences:

• important signal was missed
• affected-node reports were overlooked
• weak warnings became crisis
• hidden debt accumulated
• repair was delayed
• recurrence was misread
• dashboards replaced reality
• official memory omitted prior signals
• decisions were made from shallow context
• legitimacy shock followed ignored warnings

Accountability questions:

• What signal was available?
• Who saw it?
• Who did not?
• Was it processed or merely received?
• Was attention captured by urgency or metrics?
• Were affected-node signals reviewed?
• Were prior warnings retrieved from memory?
• Did attention limits cause delayed repair?
• Did low attention lead to misclassification?
• What attention infrastructure failed?

If attention_capacity Was Misread

Possible misread forms:

• ignored low-value noise mistaken for missed signal
• careful filtering mistaken for low attention
• delayed response mistaken for inattention when deep review was occurring
• crisis focus mistaken for neglect when triage was correct
• low signal volume mistaken for attention failure
• high signal intake mistaken for high attention
• summary reading mistaken for source blindness when summary was adequate
• automation mistaken for inattention when it is correctly bounded

Required Restoration

When attention capacity failure is found:

identify missed or overloaded signal field
→ reduce noise and nonessential load
→ restore attention budget
→ protect affected-node and weak-signal pathways
→ retrieve relevant U7 memory
→ reprocess decisions made under low attention
→ repair resulting hidden debt

If attention was asymmetrically distributed, MS-Gate should review whose signal was attended to, whose was ignored, and who carried cost from attention failure.

17) Cross-Domain Examples

Technical / Engineering

Alert volume is so high that critical warnings are ignored until outage.

Diagnostic implication: signal load exceeded attention capacity.

Operator sequence: alert triage → reduce noise → protect critical signals → U7 incident memory → recurrence monitoring.

Institutional / Governance

Complaint intake exists, but staff workload is too high to read patterns across cases.

Diagnostic implication: formal feedback exists, but attention capacity is too low for FI integrity.

Operator sequence: review backlog audit → staffing/review repair → recurrence pattern detection → affected-node validation.

AI / Algorithmic

A model has access to retrieved documents but fails to use relevant context because too much information enters the prompt.

Diagnostic implication: information access exceeded attention/context capacity.

Operator sequence: retrieval filtering → source prioritization → citation trace → answer validation.

Interaction / Relational

One person repeatedly names the same boundary issue, but the other only notices when rupture occurs.

Diagnostic implication: boundary signal is not receiving sufficient attention before crisis.

Operator sequence: ↺ reflection → explicit boundary memory → reduced load → recurrence check.

Archive / Framework Design

The project grows so fast that glossary drift, cross-link inconsistencies, and status errors are not noticed until readers become confused.

Diagnostic implication: archive signal load exceeded attention capacity.

Operator sequence: pause expansion → glossary/cross-link audit → U7 version repair → review cadence.

18) Test Protocols

1. Signal Load Test

How much signal is entering the system?

Failure signal: signal volume exceeds review capacity.

2. Attention Allocation Test

Where is attention actually going?

Failure signal: high-risk areas receive little attention.

3. Feedback Backlog Test

How much feedback is waiting unprocessed?

Failure signal: backlog age exceeds correction window.

4. Weak-Signal Test

Can low-volume but high-importance signals be noticed?

Failure signal: only loud signals move the system.

5. Affected-Node Signal Test

Are affected-node reports attended to?

Failure signal: affected-node cost is discovered late.

6. Memory Retrieval Test

Is prior signal retrieved during decisions?

Failure signal: known patterns are rediscovered.

7. Dashboard Blindness Test

Is attention captured by metrics?

Failure signal: dashboard health hides field degradation.

8. Boundary Signal Test

Are boundary strain signals noticed before breach?

Failure signal: boundary repair begins only after rupture.

9. Decision Depth Test

Does attention load collapse decision depth?

Failure signal: decisions become shallow under load.

10. Consequence Prioritization Test

Is attention allocated by consequence severity?

Failure signal: low-consequence noise crowds out high-consequence signal.

19) Anti-Patterns

• Signal received as signal processed
• Awareness as attention
• Dashboard as reality
• Alert volume as safety
• Feedback channel as feedback attention
• Low complaint rate as low issue rate
• Urgent as important
• Loud as relevant
• High-status signal as high-priority signal
• Memory stored as memory used
• Summary as sufficient by default
• Crisis attention as repair attention
• Monitoring as interpretation
• Review backlog as harmless
• Context compression as clarity
• Missed warning as surprise
• Attention exhaustion as lack of care
• Metrics as attention map
• Formal audit as actual inspection
• Silence as absence of signal

20) Spec Validation Check

• Is this truly a diagnostic, not an operator? Yes.
• Does it measure state, capacity, risk, or response rather than act directly? Yes.
• Does it map to S? Yes.
• Are U-layers specified? Yes.
• Are leading and lagging indicators separated? Yes.
• Are interpretation risks defined? Yes.
• Are operator sequencing implications clear? Yes.
• Are gate implications clear? Yes.
• Are scaling risks included? Yes.
• Are interaction implications included? Yes.
• Does it avoid new primitives? Yes.

Condensed Archive Summary

attention_capacity is the diagnostic estimate of how much meaningful signal, complexity, feedback, memory, risk, boundary strain, affected-node cost, and environmental change a system can notice, hold, prioritize, interpret, and route into repair without losing coherence or collapsing decision depth. It does not measure intelligence or intent; it measures usable attention relative to signal load and consequence severity. Low attention_capacity indicates risk of missed-signal debt, feedback backlog, weak-signal loss, affected-node invisibility, dashboard blindness, memory non-use, recurrence misrecognition, delayed repair, boundary signal loss, shallow decisions, and forced-response governance. Under low attention capacity, the system should pause expansion, reduce noise and nonessential load, triage by consequence, restore attention budget, protect affected-node and weak-signal channels, retrieve U7 memory, and avoid high-risk binding, irreversible action, deep coupling, scaling responsibility, or metric-only closure until adequate attention is restored.