WFGY/ProblemMap/Atlas/troubleshooting-atlas-router-v1.txt

AI_NOTE_START

FORMAT_NOTICE:
machine-readable first
for human-readable explanations and full documentation, use the GitHub WFGY repository

ORIGIN:
standalone routing pack derived from the GitHub WFGY repository and Problem Map 3.0 Troubleshooting Atlas

ASSET_NAME:
troubleshooting-atlas-router-v1

ROLE:
route-first pack for troubleshooting and failure classification

PRIMARY_USE:
- classify case
- identify earliest decisive failure
- choose highest honest resolution
- suggest first repair direction
- warn against likely misrepair

NOT_FULL_REPLACEMENT_FOR:
- full Atlas
- full Casebook
- full Adapter
- full repair engine
- autonomous agent
- full root-cause closure

CORE_IDEA:
route first
repair second
structural failure over topic similarity
earliest decisive failure over downstream consequence
honest coarse fit over decorative precision

REQUIRED_OUTPUT:
- primary_family
- secondary_family
- why_primary_not_secondary
- broken_invariant
- best_current_fit
- first_fix_direction
- misrepair_risk
- confidence
- evidence_sufficiency

OPTIONAL_OUTPUT:
- need_more_evidence

VALUE_SPACE:
confidence = high | medium | low
evidence_sufficiency = sufficient | partial | weak
secondary_family = family_name | none
best_current_fit = family-level | node-level | unresolved_subtype | no-fit

CORE_RULES:
1. route first; repair second
2. use structural failure, not topic words
3. prefer earliest decisive failure
4. use highest honest resolution only
5. if evidence is thin, stay coarse
6. secondary_family is not decoration
7. give first repair direction only
8. confidence must follow evidence
9. structural evidence beats vocabulary cue
10. if Atlas plausibly covers case at family-level, do not jump to no-fit

AI_NOTE_END

ROUTER_NAME:
Troubleshooting Atlas Router v1

ROUTER_ONE_LINE:
standalone route-first TXT pack for Problem Map 3.0 Troubleshooting Atlas

SECTION_1_CORE_CONTRACT

TASK:
reduce wrong-family routing
reduce wrong first-fix choice
preserve honest fit resolution

SUCCESS_CONDITION:
- primary_family honest
- broken_invariant useful
- best_current_fit not decorative
- first_fix_direction structural
- misrepair_risk real

DO_NOT:
- solve whole case
- fake precision
- use ambiguity as escape hatch
- give long speculative repair plans
- classify from rhetoric alone

QUICK_ROUTE_WORKFLOW:
1. normalize_case
2. choose_primary_family
3. test_nearest_competing_family
4. set_fit_resolution
5. set_first_fix_and_misrepair

SECTION_2_CASE_INTAKE_NORMALIZATION

NORMALIZE_CASE_FIRST:
convert raw case into structural failure description before routing

SEPARATE:
- symptom vs earliest_decisive_failure
- missing_visibility vs actual_boundary_failure
- truth_anchor_loss vs carrier_distortion
- continuity_loss vs execution_skeleton_failure
- danger_wording vs real_drift_erosion_fragmentation_overshoot
- symbolic_vocabulary vs true_representation_failure
- runtime_stress vs true_execution_closure_failure

INTAKE_QUESTIONS:
Q1: what fails first
Q2: what structural evidence supports that reading
Q3: what neighbor may absorb case too early
Q4: what missing evidence would change the cut

SECTION_3_SECONDARY_FAMILY_DISCIPLINE

USE_SECONDARY_ONLY_IF:
- neighboring pressure changes reading
- neighboring pressure matters for boundary explanation
- neighboring pressure matters for first-fix warning

DO_NOT_USE_SECONDARY_IF:
- rhetorical sophistication only
- main reading already clean
- neighboring pressure is speculative and non-material

DEFAULT:
if neighboring pressure weak -> secondary_family: none

RULE:
one honest primary > vague primary + decorative secondary

SECTION_4_TIE_BREAK_PRINCIPLES

TIE_BREAKS:
1. earliest decisive failure > later consequence
2. visible carrier failure > progression diagnosis when carrier clearly broken
3. true anchor loss > formal resemblance in world/evidence/target mismatch
4. visibility deficit > higher-order policy interpretation when path still not exposed
5. continuity-thread failure > execution-skeleton speculation unless operational dependency failure is explicit
6. explicit operational closure failure > generic continuity language when bridge/readiness/ordering/liveness evidence is concrete
7. real drift/erosion/fragmentation/overshoot > danger wording
8. if two readings remain plausible under thin evidence -> stay coarse

SECTION_5_CONFIDENCE_CALIBRATION

CONFIDENCE_HIGH_IF:
- primary cut explicit
- nearest competing family explicitly rejected
- broken invariant clear
- first_fix stable under current evidence
- evidence_sufficiency = sufficient

CONFIDENCE_MEDIUM_IF:
- primary plausible
- nearest competing family still live
or
- subtype unstable while family stable
or
- evidence_sufficiency = partial

CONFIDENCE_LOW_IF:
- only coarse family plausible
or
- decisive evidence weak
or
- competing family unresolved
or
- first_fix still provisional

CONFIDENCE_ALIGNMENT_RULE:
confidence must not exceed evidence_sufficiency

SECTION_6_SEVEN_FAMILY_OPERATIONAL_CORE

F1_NAME:
Grounding & Evidence Integrity
F1_WHAT_FAILS_FIRST:
loss of alignment with evidence anchors, truth anchors, world anchors, referents, or semantic targets
F1_BROKEN_INVARIANT:
anchor_to_claim_coupling_broken
F1_NOT:
- not primarily carrier failure
- not chosen only because answer seems vague or wrong
- not chosen from truth/meaning/realism vocabulary alone
F1_CONFUSION:
- F7
- F5
F1_TRIGGER:
earliest = evidence mismatch | target mismatch | referent drift | world-target miss | truth-anchor loss
F1_FIRST_FIX_PATTERN:
- re-grounding
- evidence verification
- world-anchor check
- target-reference audit
- proxy_vs_target separation
F1_MISREPAIR:
rewrite style or prompt tone before anchor restoration
F1_MIN_EVIDENCE:
need anchor/evidence/referent/target/world mismatch signal

F2_NAME:
Reasoning & Progression Integrity
F2_WHAT_FAILS_FIRST:
inferential path, decomposition path, recursive path, or collapse-recovery path fails before grounding, continuity, or carrier fidelity becomes decisive
F2_BROKEN_INVARIANT:
progression_continuity_broken
F2_NOT:
- not primarily descriptor/carrier failure
- not chosen from logic/reasoning/math vocabulary alone
F2_CONFUSION:
- F7
F2_TRIGGER:
carrier usable enough, but inferential path stalls, collapses, loops, fragments, or cannot recover
F2_FIRST_FIX_PATTERN:
- decomposition reset
- interpretation checkpoint
- recursive horizon fencing
- recovery scaffold
- progression guard
F2_MISREPAIR:
redesign carrier when actual first failure is progression path
F2_MIN_EVIDENCE:
need inferential path failure, progression break, recursive instability, or collapse-recovery signal

F3_NAME:
State & Continuity Integrity
F3_WHAT_FAILS_FIRST:
memory, role, ownership, continuity thread, or multi-agent interaction thread loses stability before contract or execution closure becomes primary
F3_BROKEN_INVARIANT:
state_continuity_broken
F3_NOT:
- not primarily ordering/readiness/bridge/liveness failure
- not chosen from memory/identity/agent vocabulary alone
F3_CONFUSION:
- F4
- F6
F3_TRIGGER:
earliest = role continuity | ownership continuity | memory persistence | persistence drift | interaction-thread instability
F3_FIRST_FIX_PATTERN:
- continuity restoration
- role fencing
- persistence guard
- provenance tracing
- interaction stabilization
F3_MISREPAIR:
add more instructions without restoring continuity infrastructure
F3_MIN_EVIDENCE:
need persistence break, role contamination, ownership discontinuity, continuity drift, or interaction-thread instability

F4_NAME:
Execution & Contract Integrity
F4_WHAT_FAILS_FIRST:
execution skeleton, ordering, readiness, bridge, liveness, closure path, or rule-to-action contract fails before continuity or visibility becomes primary
F4_BROKEN_INVARIANT:
execution_skeleton_closure_broken
F4_NOT:
- not primarily role/memory continuity problem
- not chosen only because runtime fails
F4_CONFUSION:
- F3
- F5
- F6
F4_TRIGGER:
earliest = ordering | readiness | bridge integrity | protocol closure | enforcement closure | deadlock | liveness
F4_FIRST_FIX_PATTERN:
- readiness audit
- ordering validation
- bridge check
- liveness watchdog
- closure-path trace
- rule-to-action restoration
F4_MISREPAIR:
improve reasoning or prompt quality before fixing runtime skeleton
F4_MIN_EVIDENCE:
need execution dependency failure, readiness failure, deadlock, bridge failure, closure failure, or liveness collapse

F5_NAME:
Observability & Diagnosability Integrity
F5_WHAT_FAILS_FIRST:
system cannot expose, trace, audit, diagnose, or anticipate relevant failure structure before deeper intervention is honest
F5_BROKEN_INVARIANT:
failure_path_visibility_broken
F5_NOT:
- not automatically correct because case sounds abstract or dangerous
- not substitute for true boundary failure
- not substitute for true grounding failure
F5_CONFUSION:
- F6
- F4
- F1
F5_TRIGGER:
earliest = missing path visibility | missing coherence visibility | missing fragility signature | missing audit route
F5_FIRST_FIX_PATTERN:
- observability insertion
- trace exposure
- audit-route uplift
- coherence visibility support
- warning horizon extension
F5_MISREPAIR:
launch higher-order intervention before exposing the relevant path
F5_MIN_EVIDENCE:
need missing traceability, missing auditability, missing coherence exposure, or warning blindness first

F6_NAME:
Boundary & Safety Integrity
F6_WHAT_FAILS_FIRST:
goal, control, incentive, collective, or regime boundary erodes before visibility deficit or execution closure becomes primary
F6_BROKEN_INVARIANT:
boundary_integrity_broken
F6_NOT:
- not automatically correct because case sounds dangerous, political, aligned, risky, or safety-related
- not chosen from rhetoric alone
F6_CONFUSION:
- F5
- F4
- F3
F6_TRIGGER:
earliest = real drift | erosion | fragmentation | capture | overshoot | unstable boundary crossing
F6_FIRST_FIX_PATTERN:
- alignment guard
- control-path audit
- incentive rebalancing
- safe-corridor monitoring
- damping and stabilization
F6_MISREPAIR:
improve observability only while boundary itself is already drifting
F6_MIN_EVIDENCE:
need real drift/erosion/fragmentation/capture/overshoot/unstable boundary-crossing evidence; danger wording alone insufficient

F7_NAME:
Representation & Localization Integrity
F7_WHAT_FAILS_FIRST:
structural carrier, symbolic shell, descriptor, layout, local anchor, or formal container fails before reasoning progression or external anchoring becomes decisive
F7_BROKEN_INVARIANT:
representation_container_fidelity_broken
F7_NOT:
- not automatically correct from symbols/logic/OCR/JSON/schema vocabulary alone
- not chosen from vocabulary alone
F7_CONFUSION:
- F1
- F2
- F5
F7_TRIGGER:
earliest = carrier distortion | descriptor inadequacy | local anchoring failure | structural shell failure
F7_FIRST_FIX_PATTERN:
- descriptor audit
- structural preservation
- local anchor repair
- layout validation
- formal adequacy check
F7_MISREPAIR:
repair reasoning chain or grounding surface while carrier remains untrustworthy
F7_MIN_EVIDENCE:
need carrier distortion, descriptor inadequacy, symbolic collapse, local anchor break, or structural shell failure

SECTION_7_MINIMAL_EVIDENCE_CHECKLIST

FAMILY_MIN_EVIDENCE_SUMMARY:
F1 -> anchor/evidence/referent/target/world mismatch
F2 -> inferential path failure/progression break/recursive instability
F3 -> continuity drift/persistence break/role contamination/ownership break
F4 -> readiness/ordering/bridge/liveness/closure/enforcement failure
F5 -> missing traceability/auditability/coherence exposure/warning visibility
F6 -> real drift/erosion/fragmentation/overshoot/capture/unstable boundary crossing
F7 -> carrier distortion/descriptor inadequacy/structural shell/local anchor failure

THIN_EVIDENCE_GATE:
if evidence thin and family plausible -> family-level
if subtype unstable -> unresolved_subtype
if competing family unresolved -> lower confidence, do not force node-level
if Atlas plausibly covers case -> prefer need_more_evidence over no-fit

SECTION_8_BOUNDARY_DECISION_MATRIX

B_F1_F7:
A=F1
B=F7
A_if=truth_anchor_loss|evidence_mismatch|referent_drift|world_target_failure first
B_if=carrier_distortion|descriptor_failure|structural_shell_failure|local_anchor_failure first
not_A_if=symbolic_or_synthetic_appearance_only_without_anchor_evidence
not_B_if=representation_words_only_without_carrier_evidence
conf_A=explicit_anchor_mismatch
conf_B=explicit_carrier_failure
tie=anchor_break>carrier_if_external_mismatch_explicit_else_coarse

B_F5_F6:
A=F5
B=F6
A_if=missing_visibility|missing_traceability|missing_auditability|warning_blindness|uninspectable_coherence first
B_if=drift|erosion|capture|fragmentation|overshoot|unstable_boundary_crossing first
not_A_if=path_already_visible_and_boundary_already_failed
not_B_if=danger_wording_only_without_real_boundary_evidence
conf_A=explicit_observability_or_audit_deficit
conf_B=explicit_drift_erosion_fragmentation_overshoot
tie=unexposed_path>F5; visible_boundary_failure>F6; otherwise coarse

B_F3_F4:
A=F3
B=F4
A_if=memory|role|ownership|continuity_thread|interaction_thread failure first
B_if=readiness|ordering|bridge|liveness|execution_closure|rule_to_action failure first
not_A_if=coordination_loss_is_really_execution_dependency_failure
not_B_if=runtime_failure_is_really_continuity_thread_loss
conf_A=explicit_continuity_evidence
conf_B=explicit_operational_dependency_evidence
tie=continuity_first>F3; execution_skeleton_first>F4; otherwise coarse

B_F2_F7:
A=F2
B=F7
A_if=carrier_usable_but_inferential_path_breaks
B_if=formal_or_symbolic_carrier_inadequate_before_reasoning_can_be_trusted
not_A_if=logic_or_math_words_only_without_path_evidence
not_B_if=formal_appearance_only_without_carrier_break
conf_A=explicit_progression_failure
conf_B=explicit_carrier_or_descriptor_failure
tie=broken_carrier>F7; broken_progression_under_usable_carrier>F2; otherwise coarse

B_F1_F5:
A=F1
B=F5
A_if=explicit_anchor_failure_first
B_if=path_too_opaque_to_establish_anchor_failure
not_A_if=answer_looks_wrong_but_anchor_break_not_exposed
not_B_if=anchor_mismatch_already_explicit
conf_A=anchor_evidence
conf_B=audit_or_visibility_deficit
tie=visible_anchor_break>F1; opaque_path>F5

B_F4_F5:
A=F4
B=F5
A_if=execution_skeleton_break_explicit
B_if=too_opaque_to_determine_execution_break
not_A_if=execution_failure_only_suspected
not_B_if=execution_dependency_failure_already_exposed
conf_A=concrete_execution_structure_evidence
conf_B=real_missing_traceability
tie=opaque_execution_path>F5; explicit_skeleton_break>F4

B_F3_F6:
A=F3
B=F6
A_if=actor|role|ownership|interaction|persistence continuity fails first
B_if=boundary_drift|fragmentation|capture|unsafe_transition fails first
not_A_if=actor_instability_language_but_real_issue_is_collective_boundary_erosion
not_B_if=danger_collective_language_but_real_issue_is_continuity
conf_A=continuity_evidence
conf_B=boundary_evidence
tie=continuity_first>F3; boundary_drift_first>F6

B_F4_F6:
A=F4
B=F6
A_if=rule_to_action|bridge|enforcement|protocol|operational_order breakdown first
B_if=boundary_erosion|legitimacy_drift|fragmentation|overshoot first
not_A_if=operational_stress_but_real_issue_is_boundary_failure
not_B_if=institutional_or_political_language_but_real_issue_is_closure_failure
conf_A=operational_closure_evidence
conf_B=real_boundary_or_regime_evidence
tie=operational_closure_first>F4; boundary_regime_failure_first>F6

SECTION_9_RESOLUTION_DISCIPLINE

FAMILY_LEVEL:
use_when=family structurally supported but node/subtree evidence too thin
note=valid_outcome_not_weakness

NODE_LEVEL:
use_when=direct_structural_evidence_for_specific_node_exists
do_not_use_when=node_name_only_sounds_close

UNRESOLVED_SUBTYPE:
use_when=family clear but subtype selection premature
note=better_than_forced_node_name

NEED_MORE_EVIDENCE:
use_when=Atlas plausibly covers case but decisive routing depends on missing evidence
do_not_use_when=missing signal would not materially change route

NO_FIT:
use_when=real_structural_mismatch_with_Atlas
do_not_use_when=family-level plausible fit still exists

DOWNGRADE_RULES:
- weak_node_evidence -> family-level
- competing_plausible_families_under_thin_evidence -> coarse
- missing_subtype_signal -> unresolved_subtype
- confidence_outpaces_evidence -> lower_confidence_then_lower_resolution

SECTION_10_FIRST_FIX_AND_MISREPAIR

FIRST_FIX_RULE:
- structural_not_cosmetic
- address_broken_invariant_first
- not_full_repair_script
- plausible_under_current_evidence
- first_move_only

MISREPAIR_RULE:
- name_tempting_wrong_first_move
- explain_why_tempting
- explain_why_it_misses_earliest_decisive_failure

PAIRING_FORMAT:
first_fix_direction = short_structural_move
misrepair_risk = tempting_wrong_move | why_tempting | why_wrong

SECTION_11_AMBIGUOUS_NOFIT_DISCIPLINE

AMBIGUOUS:
rare
use_only_if=two_or_more_family_readings_remain_genuinely_unresolved_after_structural_analysis
do_not_use_if=case_is_only_complex_or_high_abstract

NO_FIT:
rarer
use_only_if=real_structural_miss
do_not_use_if=family-level plausibility remains

COARSE_ROUTING_PREFERENCE:
if honest_family_level_fit_exists -> prefer_coarse_fit_over_decorative_uncertainty

ASK_MORE_EVIDENCE_RULE:
request_more_evidence_only_if=missing_signal_would_materially_change_route_or_fit_resolution

SECTION_12_EXEMPLAR_COVERAGE_INDEX

EX1=F1|F5
EX2=F7|F2
EX3=F5|F4
EX4=F4|none
EX5=F3|F4
EX6=F6|F5
EX7=F1|F7
EX8=F5|F6
EX9=F4|F7
EX10=F3|F5
EX11=F5|F6|thin_evidence
EX12=F1|F7|abstract_anchor

SECTION_13_GOLD_MINI_EXEMPLARS

EX1_F1_GROUNDING_FIRST:
case=answer cites retrieved chunks but chosen chunks do not support actual claim
primary=F1
secondary=F5
why=evidence-anchor mismatch arrives before opacity
broken_invariant=evidence_anchor_integrity_broken
first_fix=re_grounding|evidence_verification
misrepair=add_explanation_before_anchor_restoration

EX2_F7_CONTAINER_FIRST:
case=formal descriptor cannot preserve required reasoning structure
primary=F7
secondary=F2
why=carrier fails before progression can be trusted
broken_invariant=formal_descriptor_fidelity_broken
first_fix=descriptor_audit|formal_adequacy_validation
misrepair=improve_reasoning_steps_while_carrier_broken

EX3_F5_OBSERVABILITY_FIRST:
case=system fails repeatedly but logs do not expose failure path
primary=F5
secondary=F4
why=path invisibility arrives before confirmed execution closure failure
broken_invariant=failure_path_visibility_broken
first_fix=trace_exposure|observability_insertion
misrepair=rewrite_execution_logic_before_path_exposure

EX4_F4_EXECUTION_FIRST:
case=workflow stalls because readiness dependencies and bridge conditions never satisfied
primary=F4
secondary=none
why=execution skeleton collapse arrives before reasoning weakness
broken_invariant=readiness_and_closure_integrity_broken
first_fix=ordering_validation|bridge_checks
misrepair=improve_prompts_before_fixing_dependencies

EX5_F3_CONTINUITY_FIRST:
case=across turns system loses role ownership and stable interaction thread
primary=F3
secondary=F4
why=continuity thread loss arrives before bridge closure failure
broken_invariant=interaction_thread_continuity_broken
first_fix=role_fencing|continuity_restoration
misrepair=stronger_instructions_without_restoring_persistence

EX6_F6_BOUNDARY_FIRST:
case=control system observable but incentive drift pushes unsafe behavior
primary=F6
secondary=F5
why=boundary drift arrives before missing visibility
broken_invariant=control_boundary_integrity_broken
first_fix=control_path_audit|incentive_rebalancing
misrepair=add_observability_only_while_boundary_drifting

EX7_F1_F7_SYNTHETIC:
case=model in synthetic environment fails to extract stable truth-like anchors
primary=F1
secondary=F7
why=synthetic truth-anchor grounding fails before shell collapse
broken_invariant=synthetic_truth_anchor_grounding_broken
first_fix=truth_like_anchor_validation|referent_extraction_audit
misrepair=repair_surface_structure_before_anchor_layer

EX8_F5_F6_HIGH_ABSTRACT:
case=value_probability_knowledge system unstable but coherence structure still not inspectable
primary=F5
secondary=F6
why=missing coherence visibility arrives before proven overshoot
broken_invariant=coherence_profile_visibility_broken
first_fix=meaning_profile_tracing|coherence_visibility_uplift
misrepair=jump_to_boundary_intervention_without_structure_exposure

EX9_F4_F7_TOOL_SCHEMA:
case=tool call fails because schema/contract mismatch breaks executable invocation path
primary=F4
secondary=F7
why=operational closure fails first; schema matters because it blocks callable contract
broken_invariant=execution_contract_closure_broken
first_fix=contract_validation|tool_schema_to_runtime_check
misrepair=rewriting_reasoning_without_fixing_call_contract

EX10_F3_F5_SESSION_HANDOFF:
case=multi-turn assistant loses prior role/thread after session handoff and logs remain sparse
primary=F3
secondary=F5
why=continuity thread fails first; observability pressure is real but secondary
broken_invariant=session_continuity_broken
first_fix=persistence_guard|handoff_trace_restore
misrepair=adding_policy_text_before_restoring_state_thread

EX11_THIN_EVIDENCE_MULTI_FAMILY:
case=case sounds high-risk and unstable but evidence only shows partial opacity plus vague drift language
primary=F5
secondary=F6
why=visibility deficit is supported; boundary failure is not explicit
broken_invariant=partial_failure_path_visibility_broken
first_fix=trace_exposure|coherence_probe
misrepair=forcing_boundary_stabilization_under_thin_evidence

EX12_ABSTRACT_ANCHOR_CASE:
case=formal philosophical output looks symbolic but actual failure is mismatch with target referent
primary=F1
secondary=F7
why=anchor mismatch arrives first; formal appearance is misleading
broken_invariant=target_reference_coupling_broken
first_fix=referent_audit|target_anchor_check
misrepair=repairing_symbolic_style_before_target_alignment

SECTION_14_GOLD_OUTPUTS

OUT1_HIGH:
primary_family=F1
secondary_family=F5
why_primary_not_secondary=decisive failure is evidence-anchor mismatch; opacity exists but does not arrive first
broken_invariant=evidence_anchor_integrity_broken
best_current_fit=node-level:F1_N01 Retrieval Anchor Drift
first_fix_direction=re_grounding|chunk_to_target_trace
misrepair_risk=explanation_or_style_refinement_before_anchor_restoration
confidence=high
evidence_sufficiency=sufficient

OUT2_MEDIUM:
primary_family=F5
secondary_family=F6
why_primary_not_secondary=coherence visibility bottleneck appears first; boundary pressure may emerge later
broken_invariant=coherence_path_visibility_broken
best_current_fit=family-level
first_fix_direction=coherence_visibility_uplift|audit_route_exposure
misrepair_risk=boundary_stabilization_before_structure_is_inspectable
confidence=medium
evidence_sufficiency=partial

OUT3_LOW_NEED_MORE:
primary_family=F4
secondary_family=F3
why_primary_not_secondary=execution skeleton stress visible but continuity loss remains plausible under thin trace
broken_invariant=possible_closure_path_instability
best_current_fit=family-level
first_fix_direction=collect_more_readiness_ordering_continuity_evidence
misrepair_risk=forcing_bridge_repair_interpretation_too_early
confidence=low
evidence_sufficiency=weak
need_more_evidence=clarify_whether_ordering_bridge_failure_or_continuity_thread_loss_arrives_first

OUT4_FAMILY_ONLY:
primary_family=F7
secondary_family=F2
why_primary_not_secondary=carrier stress is clear but subtype evidence insufficient
broken_invariant=representation_carrier_integrity_weakened
best_current_fit=family-level
first_fix_direction=descriptor_and_carrier_audit
misrepair_risk=progression_first_reading_before_carrier_check
confidence=medium
evidence_sufficiency=partial

OUT5_TWO_FAMILY_COMPETITION:
primary_family=F5
secondary_family=F6
why_primary_not_secondary=both diagnosability pressure and boundary pressure exist, but current evidence supports opacity first
broken_invariant=auditability_and_coherence_visibility_bottleneck
best_current_fit=family-level
first_fix_direction=trace_exposure|coherence_visibility_uplift
misrepair_risk=using_F6_stabilization_before_boundary_failure_is_explicit
confidence=low
evidence_sufficiency=partial

OUT6_HIGH_RISK_SOUNDING_THIN:
primary_family=F6
secondary_family=F5
why_primary_not_secondary=control drift evidence is explicit, but observability pressure remains unresolved
broken_invariant=control_boundary_stability_degraded
best_current_fit=unresolved_subtype
first_fix_direction=control_path_audit|incentive_boundary_check
misrepair_risk=treating_case_as_pure_visibility_problem_after_drift_is_already_visible
confidence=medium
evidence_sufficiency=partial

SECTION_15_OVERLAY_SHORT_RULE

OBS:
use_if=visibility|traceability|auditability|interpretability pressure dominates
type=overlay_not_primary_family

SEC:
use_if=boundary_abuse|adversarial_manipulation|exploitative_misuse|attack_pressure clear
type=overlay_not_primary_family

LOC:
use_if=layout|OCR|placement|spatial_text_anchor|local_structural_alignment failure dominates
type=overlay_not_primary_family

MAIN_FAMILY_FIRST:
decide_main_family_first
overlay_refines_reading
overlay_does_not_replace_routing
overlay_does_not_become_primary_family

SECTION_16_FIT_CANDIDATE_REGISTRY

F1_FITS:
stable_enough=Retrieval Anchor Drift|Semantic Grounding Mismatch|Synthetic Truth Grounding|OOD World Grounding Failure|Train-Deploy Distribution Grounding Gap
use_with_caution=Policy-to-World Grounding Failure
family_level_only_if_thin=general_grounding_failure|evidence_anchor_mismatch|target_reference_drift
do_not_overreach=no_specific_grounding_subtype_without_clear_anchor_target_referent_world_evidence

F2_FITS:
stable_enough=Interpretation Collapse|Long-Chain Reasoning Continuity Failure|Collapse-Recovery Failure
use_with_caution=Recursive Horizon Instability|Symbolic Progression Breakdown|Philosophical Recursion Pressure
family_level_only_if_thin=reasoning_progression_failure|recursive_instability|decomposition_failure
do_not_overreach=no_named_reasoning_subtype_without_visible_inferential_pattern

F3_FITS:
stable_enough=Memory Continuity Failure|Role / Ownership Contamination|Viable State-Space Collapse|Multi-Agent Continuity Instability
use_with_caution=Ownership Continuity Break|Interaction-Thread Drift
family_level_only_if_thin=continuity_failure|persistence_break|role_thread_instability
do_not_overreach=no_specific_continuity_subtype_without_isolation_of_role_ownership_memory_interaction

F4_FITS:
stable_enough=Bootstrap Ordering Failure|Deployment Deadlock|Pre-Readiness Execution Failure|Bridge Integrity Failure|Cross-Layer Liveness Degradation
use_with_caution=Institutional Enforcement Drift|Rule-to-Action Closure Failure|Accountability Path Thinning|Fallback Realism Gap
family_level_only_if_thin=execution_skeleton_failure|closure_path_failure|operational_dependency_collapse
do_not_overreach=no_detailed_F4_fit_if_ordering_bridge_liveness_enforcement_only_guessed_from_symptom

F5_FITS:
stable_enough=Failure Path Opacity|Oversight Coverage Failure|Coherence Observability Deficit|Meaning-Profile Auditability Failure
use_with_caution=Scalable Interpretability Pressure|Early Warning Deficit|Fragility Signature Blindness|Regime-Shift Warning Delay
family_level_only_if_thin=diagnosability_failure|auditability_failure|warning_visibility_failure
do_not_overreach=no_specific_observability_subtype_without_identifiable_missing_path_or_warning_deficit

F6_FITS:
stable_enough=Alignment Boundary Drift|Control Boundary Erosion|Incentive Boundary Distortion|Incentive Misalignment Drift|Polarization Incentive Amplification
use_with_caution=Collective Boundary Drift|Collective Legitimacy Erosion|Coordination-Boundary Fragmentation|Collective Regime Overshoot
family_level_only_if_thin=boundary_drift|control_drift|incentive_erosion|regime_instability
do_not_overreach=no_specific_F6_subtype_without_real_drift_erosion_fragmentation_capture_overshoot_evidence

F7_FITS:
stable_enough=Symbolic Representation Fidelity Failure|Logic Descriptor Fidelity Failure|Formal Container Adequacy Failure|Localization and Layout Anchoring Failure
use_with_caution=Explanation Fidelity Distortion|Synthetic Structure Fidelity Failure|Hierarchical Composition Drift|Constraint-Skeleton Distortion
family_level_only_if_thin=representation_failure|descriptor_failure|carrier_distortion|structural_shell_instability
do_not_overreach=no_specific_F7_subtype_without_visible_carrier_descriptor_layout_shell_failure

SECTION_17_AUDIT_APPENDIX

AUDIT_STRUCTURE:
- product_identity_and_scope
- core_routing_contract
- case_intake_normalization
- secondary_family_discipline
- tie_break_principles
- confidence_calibration
- seven_family_operational_core
- minimal_evidence_checklist
- boundary_decision_matrix
- resolution_discipline
- first_fix_and_misrepair
- ambiguous_no_fit_discipline
- exemplar_coverage_index
- gold_mini_exemplars
- gold_outputs
- overlay_rule
- fit_candidate_registry
- audit_appendix
- pressure_tests
- release_gate
- freeze_note

AUDIT_FAMILY_OPERATIONALIZATION:
each_family_must_have=
what_fails_first|
broken_invariant|
what_this_family_is_not|
confusion_family|
boundary_trigger|
first_fix_pattern|
misrepair_trap|
minimum_evidence

AUDIT_BOUNDARY_COVERAGE:
required_boundaries=
F1/F7|F5/F6|F3/F4|F2/F7|F1/F5|F4/F5|F3/F6|F4/F6
each_boundary_must_have=
choose_A|
choose_B|
not_A_too_early|
not_B_too_early|
evidence_threshold|
default_tie

AUDIT_RESOLUTION_COVERAGE:
must_define=
family-level|
node-level|
unresolved_subtype|
need_more_evidence|
no-fit|
downgrade_rules|
thin_evidence_gate

AUDIT_MISREPAIR_COVERAGE:
every_family_must_define_at_least_one_common_misrepair_trap

AUDIT_OUTPUT_COVERAGE:
required_fields_supported_by_examples=
primary_family|
secondary_family|
why_primary_not_secondary|
broken_invariant|
best_current_fit|
first_fix_direction|
misrepair_risk|
confidence|
evidence_sufficiency

SECTION_18_NO_LAZINESS_AUDIT

FAIL_IF_TOPIC_SHORTCUT:
- safety_words_auto_F6
- OCR_logic_schema_words_auto_F7
- memory_words_auto_F3
- execution_words_auto_F4
- abstract_language_auto_F5

FAIL_IF_DECORATIVE_SECONDARY:
secondary_family_used_only_to_sound_intelligent

FAIL_IF_DECORATIVE_NODE_FIT:
node_level_from_thematic_resemblance_without_direct_structural_evidence

FAIL_IF_AMBIGUITY_ESCAPE:
ambiguous_or_no_fit_used_when_honest_coarse_fit_exists

FAIL_IF_COSMETIC_FIX_FIRST:
first_fix_direction_drifted_to_style_prompt_surface_cleanup_before_broken_invariant

SECTION_19_PRESSURE_TESTS

TEST_1_GROUNDING_VS_CONTAINER:
can_router_distinguish_external_anchor_loss_from_structural_carrier_failure_without_collapsing_both_into_F1_or_F7

TEST_2_OBSERVABILITY_VS_BOUNDARY:
can_router_avoid_choosing_F6_too_early_when_first_bottleneck_is_missing_visibility_or_coherence_exposure

TEST_3_CONTINUITY_VS_EXECUTION:
can_router_distinguish_role_ownership_persistence_drift_from_ordering_readiness_bridge_closure_failure

TEST_4_FORMAL_LOOKING_BUT_GROUNDING_FIRST:
can_router_avoid_overreading_F7_when_case_looks_formal_but_first_real_failure_is_truth_anchor_mismatch

TEST_5_HIGH_ABSTRACT_BUT_ROUTABLE:
can_router_avoid_ambiguous_or_no_fit_too_early_on_high_abstract_cases_still_plausibly_routable_at_family_level

TEST_6_MODE_STABILITY:
same_case_under_strict_teaching_repair_first_compact_should_preserve_same_primary_family_and_same_broken_invariant_reading
explanation_depth_may_change
primary_routing_should_not_drift_casually

SECTION_20_RELEASE_GATE

PASS_IF:
1. all_7_families_operationalized
2. all_8_required_boundaries_decisionized
3. best_current_fit_resolution_rules_explicit
4. fit_candidate_registry_present
5. first_fix_and_misrepair_paired
6. at_least_12_mini_exemplars_present
7. at_least_6_gold_outputs_present
8. overlay_rule_present
9. audit_appendix_present
10. pressure_tests_present

FAIL_IF:
- family_names_without_operational_definitions
- decorative_node_fit_common
- F6_or_F7_absorb_cases_from_topic_words_alone
- ambiguous_or_no_fit_used_as_uncertainty_shortcut
- first_fix_direction_cosmetic
- output_fields_exist_without_exemplar_support

FREEZE_STATEMENT:
if_pass_and_not_fail -> troubleshooting-atlas-router-v1 may be treated as operational_freeze and first_formal_standalone_routing_pack

SECTION_21_FREEZE_NOTE

WHAT_THIS_ROUTER_IS:
first formal standalone route-first TXT pack derived from Problem Map 3.0 Troubleshooting Atlas

WHAT_IT_DOES:
- classify failures
- cut key family boundaries
- identify broken invariants
- choose honest fit resolution
- suggest first repair direction
- reduce likely misrepair

WHAT_IT_DOES_NOT_REPLACE:
- full Atlas
- full Casebook
- full Adapter
- full repair layer
- deeper WFGY exploration

FROZEN_ENOUGH_NOW:
- product identity
- TXT delivery form
- route-first output contract
- seven-family quick map
- high-value boundary layer
- confidence and evidence discipline
- first-fix restraint
- exemplar-backed routing surface

PATCHABLE_LATER:
- wording compression
- stronger exemplars
- broader domain coverage
- stronger deployment variants
- tighter future patch alignment

UNSUPPORTED_SCOPE_NOTE:
current exemplar coverage is strongest in troubleshooting-adjacent AI failures, retrieval-like grounding, control, multi-agent continuity, execution closure, and diagnosability-heavy cases
broader generation-first, dialogue-native, and code-synthesis-native cases remain patchable expansion zones

CROSS_LLM_VALIDATION_HOOK:
empirical cross-LLM validation remains patchable
future evaluation may compare routing stability, fit honesty, and first-fix quality across multiple model families

OFFICIAL_ONE_LINE:
Troubleshooting Atlas Router v1 is the first formal route-first TXT pack derived from Problem Map 3.0 Troubleshooting Atlas. It helps AI systems classify failures, cut key family boundaries, identify broken invariants, and suggest the first repair direction without overclaiming full repair closure.