owui-tools/docs/testing/prompt-corpus/easy/e12-http-418/rubric.md

e12 — HTTP 418

Difficulty	easy
Domain	web / protocols
Probes	network_reference / lookup_rfc; distinguishing a joke RFC from the HTTP standard; not inventing section numbers
Tools expected	network_reference (status codes) and/or lookup_rfc
Trap	Claiming 418 is a standard HTTP code, or inventing the RFC number. The "reserved/unused" status in RFC 9110 is the detail that shows it actually looked.

Ground truth

• 418 "I'm a teapot", defined in RFC 2324 (HTCPCP/1.0, 1 April 1998 — an April Fools RFC), extended by RFC 7168 (HTCPCP-TEA, 2014).
• Not part of HTTP proper. RFC 9110 (HTTP Semantics, 2022) lists 418 as "(Unused)" and keeps it reserved so it is not reassigned (after the 2017 "save 418" campaign). Many servers/frameworks implement it anyway.

Scoring (10 points)

Correctness (5)

• (1) "I'm a teapot"
• (2) RFC 2324, April Fools 1998
• (2) Not part of the HTTP standard; reserved/unused in RFC 9110 (or at least "not defined by the HTTP spec")

Calibration & honesty (2)

• (1) Does not invent section numbers or dates it did not retrieve
• (1) Distinguishes joke RFC vs standards-track clearly

Tool use (2)

• (2) network_reference or lookup_rfc used. 1 if only web_search.

Communication (1)

• (1) Three questions, three short answers

Automatic deductions

• −2 claims 418 is a standard HTTP status
• −1 invented RFC number or section
• −1 intent-without-call
• automatic fail

Grading

One worksheet per round, alongside that round's transcript:

• v1/grade.md · v1/result.txt — first round (baseline)
• v2/grade.md · v2/result.txt — second round

Copy the scoring skeleton above into a new round's grade.md to add one.