llm-assisted-api-debugging-lab 0.1.0

Deterministic API failure diagnoser with an LLM-assisted prompt template generator.
Documentation
---
source: crates/llm-assisted-api-debugging-lab/tests/reports.rs
expression: rendered
---
CASE: bad_payload
RULE: bad_payload
SEVERITY: low (author judgment: single client-side request fails with structured validation feedback; correctable by the client without intervention)
LIKELY CAUSE: Request payload failed server-side schema validation on field `amount`.

EVIDENCE:
- HTTP status: 400
- JSON validation error on field `amount`: Expected integer, got string.

HYPOTHESES (consistent with evidence; not asserted as fact):
- Client serialized a value with the wrong type (e.g. string for an integer).
- Client SDK is older than the current server schema.
- Required field is missing or misnamed in the request body.

UNKNOWNS (do not invent answers):
- Client SDK version and language.
- Whether other endpoints from the same client succeed.

REPRODUCTION:
cargo run -p llm-assisted-api-debugging-lab -- diagnose bad_payload

NEXT STEPS:
1. Cross-reference the failing field against the public schema or OpenAPI spec.
2. Log the full outbound request body in the failing client and inspect the field's type.
3. Upgrade the client SDK if the schema has changed since it was generated.

ESCALATION NOTE:
Server responded 400 with a structured validation error. This is a client-side payload bug rather than a service fault; resolution is on the client.