Actor definition from requirements analysis.
Optionalcontext_Step 1.25: Actor/interface/database context verification.
Correction-specific ledger that records the AutoBeAnalyze actor, Interface authorization/controller contract, real database model/field facts, and payload/decorator/provider naming facts that the corrected code preserves.
Optional for backward compatibility with histories recorded before this ledger was introduced.
Optionaldatabase_Step 2.25: Corrected database mapping ledger.
Names the exact Prisma model(s), primary/FK/relation fields, session fields, and soft-delete/expiration predicates used after correction.
Optional for backward compatibility with older histories.
NestJS parameter decorator for route protection.
Step 1: TypeScript compilation error analysis and diagnosis.
AI identifies and categorizes all compilation errors (type mismatches, import issues, syntax errors) by component (providers/decorator/payload). Lists specific error messages with their locations and types for systematic troubleshooting.
JWT token payload structure.
OptionalplanPlanning ledger captured from the initial authorization write.
Optional for backward compatibility with histories recorded before the authorization writer persisted its CoT compliance ledger.
Authorization provider function (JWT verification, role validation).
A first-class AutoBeRealizeFunction (type: "authorization") so it
rides the shared queryIR write/correct/materialize machinery. Its content
always holds the materialized (Prisma-rendered) TypeScript.
OptionalresolutionsStep 1.5: Per-diagnostic resolution ledger.
One disposition per diagnostic shown to the correction agent — either a concrete fix applied by the corrected components, or an explicit escalation when the error cannot be genuinely fixed at this layer.
Optional for backward compatibility with histories recorded before the ledger was introduced (2026-06-11).
Optionalsecurity_Step 2.5: Corrected security strategy ledger.
Confirms JWT verification, role discriminator checks, typia-backed payload narrowing, database enrollment validation, and revoked/deleted/expired checks after correction.
Optional for backward compatibility with older histories.
Step 2: Solution guidance and fix recommendations.
AI provides clear, actionable instructions on how to resolve each identified error. Includes specific steps like "add property X to interface Y", "update import path from A to B", or "change type from C to D". Focus on guidance rather than generating complete code implementations.
Step 0: Systematic correction analysis.
The correction agent must inventory every diagnostic, identify the root cause, verify actor/interface/database context, and plan a security-preserving fix before emitting corrected provider, decorator, and payload code.
Author
Michael