How the scan was classified.

“Accepted” entries belong to the mathematical catalog. “Rejected” means the extracted span is not a complete mathematical record; it is still preserved here and in its full source. Where an adjacent source passage supplies the complete expression, the recovered replacement is linked.

Download every scan decision ↗

cookie is Secure, Path=/, and has no Domain, so it cannot be planted by a sibling subdomain. The

goes through the authenticated admin ResetMFA path. The database WHERE status = 'pending' guard

p = r.values[r.i-1]

p = count

p = 37

(dest[0].(*uuid.UUID)) = target

(dest[1].(*string)) = "user@example.gov"

(dest[2].(*string)) = string(security.RoleMember)

(dest[3].(*string)) = "active"

(dest[0].(*uuid.UUID)) = uuid.New()

1234567.89 + 2000000.10 = 3234567.99 exactly; float64 addition would drift off the last cent.

demandedTotal = 20, gap unmet = 5 => 25.00 points of overall coverage.

Rounding: unmet 1 of demanded 3 => 33.33.

WCC composite weights (DISTINCT from the engine's per-candidate weights). Sum == 1.00 (tested).

Proposal-ready = green slate members on slates tied to an OPEN (or standalone) opportunity,

+ required level. Banner staffing_gaps == sum of these (incl Uncategorized) so they reconcile [COV-3].

drivers computes the four "inventory coverage" drivers (a requirement is COVERED iff >=1 qualified

male 8/10 (rate 0.80), female 4/10 (rate 0.40). Female ratio = 0.40/0.80 = 0.50, below 0.80, so

male 8/10 (0.80), female 7/10 (0.70). Ratio 0.70/0.80 = 0.875, at or above 0.80, not flagged.

as of day is still effective, matching the Command Center's `clearance_expires_at >= as_of::date`

asOf carries a wall-clock time (mimics s.clock()=time.Now()). A clearance expiring on the SAME

civil day must stay Active so this agrees with the WCC's `>= as_of::date` boundary [ENGINE-1/2].

Band cut points (CLAUDE.md): Green >= 90, Yellow 75–89.99, Red < 75. Inclusive lower bounds.

injected semantic matcher. It always returns a SkillMatch row (Matched=false when nothing met).

SkillMatch entry with Matched==true. It matches by RequiredSkillID when set, else by

Build a lookup from the matched slice: id (when non-empty) and name (lowercased) => matched.

available + exact location = 100

available + relocatable (wrong location) = (50 + 100)/2 = 75

unavailable + exact location = (100 + 0)/2 = 50

remote allowed + remote candidate = 100 location, available = 100

Skills 50. Total 25+12.5+15+15+10+10 = 87.50, Yellow, one mandatory skill gap (Go).

Total 0+12.5+0+0+6+7.5 = 26.00, Red. Gaps: clearance, mandatory Go, Security+.

ClearanceStatus gates the clearance: only an Active clearance can score. Expired/Unknown = 0.

On an exact id/name match it returns Exact=true, Similarity=1.0. ok is false when nothing

(Red=0, Yellow=1, Green=2).

FP = other rows scored into this band.

FN = this band's cases that were scored into other bands.

TS/SCI active, all required skills, cert, customer, experience, on site, available => Green.

Clearance 100*0.25=25, Skills 100*0.25=25, Certs 100*0.15=15, Customer 100*0.15=15,

Exp 100*0.10=10, Location (100+100)/2*0.10=10. Total=100.00 => Green.

Missing mandatory Machine Learning. Skills 50%. Clearance 100*0.25=25, Skills 50*0.25=12.5,

Certs 100*0.15=15, Customer 100*0.15=15, Exp 100*0.10=10, Location 100*0.10=10. Total=87.50 => Yellow.

Expired clearance => clearance 0. All other dimensions are perfect.

Total = 0+25+15+15+10+10 = 75.00. bandFor(75.00) = Yellow, but v1.4.0 clearance gap cap

fires (expired clearance zeroes the component; required level TS/SCI >= ClearanceConfidential).

no experience with the required customer agency. The engine scores: clearance 100*0.25=25,

skills 100*0.25=25 (both mandatory met), certs 100*0.15=15, customer 0*0.15=0, exp 100*0.10=10,

location 100*0.10=10. Total = 85.00 => Yellow by the engine. No clearance gap (clearance is fine).

produces Red via the clearance gap cap (expired clearance => clearance 0, required level

TS/SCI >= Confidential, so the cap fires). This is now a CORRECT classification, not a

Green precision = 2 / (2+1) = 0.666...

Green recall = 2 / 2 = 1.0 (all expected Green are scored Green).

Red recall = 0 / 1 = 0.0 (the one expected Red was scored Green).

nameToID:  "sigint"   => sigintID

"skill b"  => skillBID

"skill a"  => skillAID

aliasToID: "signals intelligence" => sigintID

rels:      skillAID => [{toID: skillBID, relation: "substitutes_for", strength: 0.80}]

AliasMatched=true to distinguish this path from a semantic cosine match.

Skills component: 1 of 1 required met (the alias resolved) => 100.

The matched skill record must carry AliasMatched=true and Similarity=1.0. The engine

0 years req, no location req, candidate is available). Total = 100.00.

skill fraction is ratio(1,1)=100. The partial credit (0.80) lives in the SkillMatch.Similarity

Skills component: ratio(1,1) = 100. The relationship strength is in the SkillMatch record,

The matched skill record must carry Similarity=0.80, Exact=false, AliasMatched=false.

skills, which resolved to 100). Total = 100.00.

All six dimensions score 100. Total = 100.00.

The total is: 0*0.25 + 100*0.25 + 100*0.15 + 100*0.15 + 100*0.10 + 100*0.10 = 75.00.

(RequiredLevel TS/SCI >= ClearanceConfidential, clearanceScore == 0). The band cap lowers

Total = 0*0.25 + 100*0.25 + 100*0.15 + 100*0.15 + 100*0.10 + 100*0.10 = 75.00.