Define the question before collecting material

The fixed synthetic job asks for a workshop content brief aimed at independent consultants. The offer is a two-hour planning session, and the brief may use confirmed details only.

The source set must support four decisions:

  • who the brief serves;
  • which offer it describes;
  • which constraint controls the prose;
  • which schedule and price questions remain unanswered.

Material about other services or audiences may matter elsewhere, but it does not help answer this job.

Separate supplied facts from working assumptions

The synthetic source note confirms three items:

  1. The audience is independent consultants.
  2. The offer is a two-hour planning session.
  3. The brief must use confirmed details only.

No schedule or price appears in the source. The first content brief adds “Saturday sessions are available.” That sentence sounds plausible, but the source cannot support it. The source library makes the gap visible before the statement reaches the corrected brief.

Record origin before synthesis

Each source entry should identify its title, origin, date, scope, and synthetic or factual status. Those fields help a later reader distinguish evidence from a demonstration.

W3C PROV-O provides a formal vocabulary for provenance relationships. A small library can use the principle without reproducing the full model: identify the item, its source, and the activity that transformed it.

Retrieve a narrow packet for the task

Select the three notes that answer the brief question. Record why each was included and why nearby material was omitted. When the sources do not support a requested claim, return a missing-evidence note instead of filling the space with a guess.

The NIST AI Risk Management Framework describes risk work as an ongoing set of governance, mapping, measurement, and management activities. For a content brief, the practical lesson is modest: define context and consequences before treating an output as usable.

Review the packet before writing

Ask four questions:

  1. Does every planned factual claim point to a selected source?
  2. Did old output slip into the source set as if it were evidence?
  3. Are exclusions stated in plain language?
  4. Can the packet say it lacks an answer?

If any answer is unclear, revise the source set before drafting the brief.

Fit and wrong fit

This method helps repeated research, content, support, and planning jobs. It adds little value when the task needs no factual grounding. It is also insufficient by itself for specialized legal, medical, or financial decisions that require qualified review.