Pack the right inputs

For turn a verified meal-prep class brief into a truthful two-variant email test, assemble only what changes the answer: approved product facts, audience problem, channel, word limit, brand voice, prohibited claims, CTA, and test hypothesis. Label each item by authority and date. A source-of-truth document outranks a memory-based note; a current error log outranks a description of last month's behavior. State conflicts instead of letting the model blend them.

Use a four-part packet: task, evidence, constraints, and output contract. Put untrusted content inside clear delimiters and say that it is data, not instruction. Include representative examples, especially one normal case and one boundary case. Omit irrelevant history; excess context can hide the one line that controls the result.

Use only these facts: two-hour in-person class; ingredients included; participants leave with five recipes. Audience: busy cooking beginners. Draft a subject line and email under 140 words with one sign-up CTA. Do not claim savings, health outcomes, superiority, scarcity, or testimonials. Then make variant B by changing only the angle from convenience to confidence.

Before sending, annotate the packet. Mark which values are verified, which are illustrative, and which are unknown. If a screenshot is involved, transcribe critical small text. If structured data is involved, include headers, units, software version, and null behavior. If creative material is involved, record ownership and permitted use. This is how context becomes operational rather than decorative.

A useful response would look like this: Two compact emails with the same offer and CTA; each claim maps to the brief, while the changed angle is explicitly labeled as the sole test variable. That description is intentionally observable. “Looks good” is not acceptance. The operator must trace every factual phrase to the approved brief, check link and disclosure requirements, review accessibility and brand voice, then measure the predeclared conversion event. Keep the source material beside the draft so review means comparison, not memory.

The privacy boundary is specific: never paste customer lists, personal profiles, unpublished campaign data, or sensitive targeting attributes into an unapproved model. Create the smallest synthetic example that preserves the problem. Replace names and identifiers consistently so relationships remain testable. Redaction is not merely drawing a box: crop surrounding notifications, remove metadata where relevant, and check that hidden sheets, comments, or revision history are not included.

Poor packets lead to predictable failures: fabricated urgency; fake statistics; multiple variables changed; vague CTA; stereotyped audience language; optimization without a hypothesis. Another common failure is silently changing the source packet mid-run. Save a version or hash of the inputs beside the output, especially when another person will reproduce the work.

Rate a packet from zero to two on six dimensions: relevance, authority, recency, completeness, privacy, and reproducibility. A score below two on authority or privacy blocks the run. A low completeness score does not invite invention; it creates a question for the owner.