CAMS RAW SCORER: COUNTRIES v1.2-OPT  (Backstory deep-time adapter DT-1)

Score each society below at its anchor year only. One block of eight rows per society.

| Society | Year | Window and scope |
| B01_Longshan_YellowRiver | -2300 | Longshan horizon, middle and lower Yellow River incl. Taosi, c. 2500-2000 BCE |
| B02_Shang_Anyang | -1200 | Late Shang at Anyang, Wu Ding to Di Yi era |
| B03_Harappan_Indus | -2300 | Mature Harappan cities, c. 2600-1900 BCE |
| B04_OldKingdom_Egypt | -2500 | Old Kingdom Egypt, late 4th to early 5th dynasty |
| B05_FirstIntermediate_Egypt | -2150 | First Intermediate Period Egypt |
| B06_OldBabylonian_Mesopotamia | -1760 | Babylon under Hammurabi |
| B07_Minoan_Crete | -1600 | Neopalatial Crete |
| B08_Mycenaean_Greece | -1250 | Mycenaean palatial Greece, LH IIIB |
| B09_LBA_Collapse_Aegean | -1175 | Aegean after the palace destructions, early LH IIIC |
| B10_Aboriginal_Australia | 1700 | Aboriginal Australia before contact, continent-wide typical patterns |
| B11_Maori_Aotearoa | 1500 | Māori Aotearoa, after settlement and moa decline, pā building |
| B12_Xiongnu_Steppe | -170 | Xiongnu confederation under Modu's successors |

## DEEP-TIME ADDENDUM (DT-1)
- Year is a window anchor, not a calendar year-end. Score the typical institutional condition of the stated window around the anchor.
- The t-1 baseline rule does not apply; each society is scored on its own evidence.
- Archaeological, epigraphic and textual indicators all count as functional evidence (e.g. storage capacity, fortification, settlement hierarchy, skeletal stress, administrative tablets, trade goods).
- Node definitions are functional. Where a modern example (universities, courts, banks) has no ancient form, score the institution performing that function (priesthood, scribes, merchants). Where nothing performs it, follow the evidence gate.
- For societies without a state, score the institutions that perform each function at the scale of the society; do not penalise the absence of state forms as such.

## OUTPUT CONTRACT
Return ONLY CSV. No prose. No headers beyond the schema line. No markdown.
No evidence notes. No Node Value. No Bond Strength. No SHI. No confidence
labels. No interpolation. No smoothing.

## PROHIBITED → REQUIRED
Explain or justify scores          → Emit CSV rows only.
Narrativise or diagnose            → Score demonstrated function vs the node definition.
Interpolate or smooth across years → Use NA for missing evidence; never copy prior-year values.
Compute derived metrics            → Output raw C,K,S,A only.
Score from sentiment or reputation → Score from concrete functional indicators per node.
Treat one headline as all-node evidence → Evaluate each node against its own evidence.

## DATA USE
Use corpus knowledge first. Do not cite sources.
If evidence is genuinely insufficient for a specific score, output NA for that
score only.

## SOURCE POSTURE
Score demonstrated institutional function. Where the available record is
dominated by one vantage — a rival state's chronicle, a conqueror's account, a
later tradition, or a coloniser's ethnography — that record establishes what was
reported, not automatically what functioned. Adversarial or celebratory framing
is not itself functional evidence.

This does not license discounting inconvenient evidence. It requires that the
indicator be functional rather than evaluative. If only evaluative
characterisation is available for a node, the gate has not been satisfied.

## EVIDENCE GATE (per node, before scoring)
Confirm ALL three before any non-NA score:
  [ ] One concrete indicator of functional performance observed.
  [ ] One concrete indicator of strain, failure, or limitation observed.
  [ ] Mechanism connecting that evidence to THIS node's specific function identified.

If any box is unchecked → score NA for all four dimensions of this node.

## UPSTREAM CONSTRAINTS (apply during generation, not after)
1. Score below 5 or above 8 requires explicit functional evidence. Controversy,
   criticism, or popularity do not suffice for <5. Absence of problems does not
   suffice for >8.
2. A single event must not automatically move all eight nodes. Each node moves
   on its own evidence.

## NODES (canonical order: Helm, Shield, Lore, Stewards, Craft, Hands, Archive, Flow)

HELM
  Function: Executive coordination, state strategy, political centre, governing command.
  Boundary: Capacity is demonstrated execution, not mandate or claimed authority.
  Common error: Scoring command from monuments' propaganda rather than execution.

SHIELD
  Function: Military, police, guards, border control, internal order, territorial defence.
  Boundary: Coercive and protective capacity. Equipment lists alone do not
    demonstrate operational coordination.
  Common error: Conflating spending or display with functional readiness.

LORE
  Function: Religion, education, knowledge institutions, public meaning,
    legitimacy production, shared understanding.
  Boundary: Production, validation and circulation of knowledge and meaning.
    Contestation may reduce Coherence and raise Stress without eliminating
    Capacity or Abstraction.
  Common error: Scoring ideological alignment as Capacity.

STEWARDS
  Function: Landholders, owners of stores and herds, fiscal elites, resource
    allocation, ownership power.
  Boundary: Owners and managers of stored resources and long-lived assets
    specifically. NOT administration generally — that is Helm or Archive.
  Common error: Substituting bureaucratic administration for asset-holding stewardship.

CRAFT
  Function: Skilled trades, builders, metalworkers, specialised production.
  Boundary: Productive transformation and technical competence. Wealth display
    does NOT demonstrate Craft strength.
  Common error: Scoring market value of holdings as productive capacity.

HANDS
  Function: Mass labour, agricultural labour, corvée, bodily mobilisation,
    demographic work capacity.
  Boundary: Ample labour raises Capacity but does NOT cancel subsistence,
    coercion or precarity Stress. Both may be true simultaneously.
  Common error: Using output alone and ignoring the condition of the labourers.

ARCHIVE
  Function: Records, law, scribes, administration, institutional memory,
    continuity across time (written or oral).
  Boundary: Preservation, retrieval and transmission of institutional memory.
  Common error: Using later narrative traditions as direct measures of record function.

FLOW
  Function: Exchange, trade, merchants, logistics, harbours, money or value
    tokens, circulation of goods.
  Boundary: Whether exchange and circulation function, not whether they were fair.
  Common error: Scoring policy intent rather than operational circulation.

## METRICS (integer 1-10 or NA)

COHERENCE — internal alignment and coordination clarity
  9-10  Exceptionally unified or seamless. Requires historically unusual evidence.
  7-8   Integrated, strongly aligned.
  5-6   Functional, coordinated despite tension.
  3-4   Divided, factional, inconsistent.
  1-2   Fragmented, paralysed, openly conflicting.

CAPACITY — demonstrated ability to perform the node function
  9-10  Dominant or exceptional. Requires comprehensive, historically unusual evidence.
  7-8   Strong, effective, resilient.
  5-6   Adequate, performs core function.
  3-4   Weak, insufficient, unreliable.
  1-2   Collapsed or unable to function.

STRESS — rate of breakdown or entropy production
  9-10  Rupture, active collapse or severe breakdown.
  7-8   Strained, visible degradation.
  5-6   Pressured, stretched but functioning.
  3-4   Stable, challenges managed.
  1-2   Thriving, low breakdown, strengthening.

ABSTRACTION — operational sophistication, learning capacity, modelling depth
  9-10  Frontier sophistication for its era or transformative learning.
  7-8   Advanced modelling, adaptation, institutional learning.
  5-6   Functional learning and procedural sophistication.
  3-4   Basic systems, limited adaptation.
  1-2   Reactive, little learning or continuity.

## TRANSITIONS
Do not automatically score transitions as crises. Score the observed
coordination dynamics: did the node maintain coordination; could it perform its
function; was change managed or chaotic? Managed transition → moderate stress.
Chaotic rupture → high stress.

## SPECIAL CASES
BCE years are negative integers. Fragmented polities: score the dominant or most
institutionally coherent actor performing each function; still emit one row per node.

## INDEPENDENT PASS INTEGRITY
This is one isolated scoring pass. Score entirely on your own judgement against
this rubric. Do not adjust toward any expected result.

## OUTPUT FORMAT
First line exactly:

Society,Year,Node,Coherence,Capacity,Stress,Abstraction

Then 96 rows: societies in the order listed, nodes in the order
Helm, Shield, Lore, Stewards, Craft, Hands, Archive, Flow.
All scores integers 1-10 or NA. No decimals. No blanks. No extra columns.
