Fix detected mismatches in AI-generated code: rename drifted functions, correct field names, add missing parameters, and resolve contract violations. Uses `camouf_suggest_fix` to get precise fix instructions, then applies edits. Use after running validate when ERROR violations are found.
You have access to camouf_validate and camouf_suggest_fix MCP tools. Use them
to detect and fix architecture violations caused by AI-generated code.
camouf:validate found ERROR violationsFollow this exact loop (maximum 3 iterations):
camouf_validate to get all current violationscamouf_suggest_fix with:
ruleId: the rule that flagged the violationfile: the file containing the violationline: the line numberviolationId: the violation ID (if available)camouf_validate again to confirm fixes workedRepeat steps 1-5 for any remaining violations. Stop after 3 iterations or when zero ERROR violations remain.
// Before (wrong): getUser(userId)
// After (fixed): getUserById(userId)
Rename the function call to match the canonical exported name.
// Before (wrong): cancelOrder(orderId)
// After (fixed): cancelOrder(orderId, reason)
Read the surrounding code to determine the correct value for the missing parameter. Check request bodies, user inputs, or use a sensible default.
// Before (wrong): user.userEmail
// After (fixed): user.email
Replace the incorrect field name with the one defined in the shared type.
// Before (wrong): import { validate } from './helpers'
// After (fixed): import { validateInput } from './utils/validation'
Find the actual module and function name, update the import path and binding.