Methodology Chapter 01 of 08
How a Story becomes a summary.
Two things you read in Mosaic News are written by AI: the summary on a Story, which is one event as covered by many outlets, and the narrative on a Big Picture, which gathers related Stories into a longer arc. How both are assembled is the Story matching chapter. Both are labeled as AI-generated, and built only from the articles placed in front of the model.
One Story, read from every side.
One Story shows its Center, Left, and Right summaries, each drawn only from the sources carrying that lean label.
Trump Seeks $87.6B for Iran War Costs
✦ Summary updated June 25, 2026
Summary
The Trump administration requested $87.6 billion in supplemental funding from Congress on Wednesday, with $67 billion designated for the Department of Defense to cover costs from the Iran war (Operation Epic Fury). The request includes $21 billion for munitions, $17.3 billion for operational costs, and $12.1 billion for classified programs. The package also allocates $11.1 billion for U.S. farmers and $1.4 billion for Ebola response in Africa. The funding request came one day after Congress passed a war powers resolution to limit Trump's military authority against Iran. Democratic lawmakers signaled opposition, with Senator Patty Murray calling it a “disastrous war of choice,” while Republican support remained mixed.
✦ This summary is AI-generated and may contain errors or not fully capture all perspectives. Always verify important facts and refer to the original articles for accurate reporting. How we use AI →
What the model is handed
Writing a summary is one self-contained question, asked once and answered once. Mosaic News currently asks Claude Haiku 4.5, a model made by Anthropic. The question has three parts and nothing else in it.
The instructions
A written rulebook telling the model what to produce, which evidence it may use, and what to leave empty when the evidence is not there. You can read it in full further down this page.
The evidence
The text of the articles themselves. For a Big Picture, it is the summaries of the Stories inside it. Each one is labeled with its publisher, that publisher's lean rating, and when it was published.
The required fields
The answer has to come back as a fixed set of fields: a title, an account of what happened, and how left-leaning and right-leaning outlets framed it. There is nowhere to write anything else.
That is the whole of it. The model cannot browse the web, it does not remember the question asked before this one, and it is never told who you are or what you have been reading.
How the reading list is chosen
A summary can only be as fair as what it was allowed to read, so the reading list is decided before the model writes anything, by rules that do not change from one Story to the next.
- At most ten articles are read for one Story summary.
- Left, center, and right coverage each get reserved places whenever that coverage exists.
- Within each of those groups, outlets with stronger factual-reporting and credibility ratings come first.
- Newer articles are preferred, and so are articles that add something the list does not already have.
- When one wire story is reprinted word for word across many sites, it counts once. Five reprints cannot take five places.
- Only full article text is used. An article that could not be retrieved in full is not sent.
Factual-reporting and credibility ratings come from Media Bias/Fact Check, the same ratings shown on every publisher's report card in the app. MBFC publishes how it produces them at mediabiasfactcheck.com/methodology.
The rules it is held to
When a language model runs short of information, it can fill the space with text that reads well and is not true. That risk cannot be reduced to zero, so the instructions work against it. The model may not draw on anything it learned in training, every claim has to trace back to one of the articles it was given, and an honest empty field is always the better answer.
Base your analysis ONLY on the article texts provided below. Do NOT use any knowledge from your training data.
The first line of the evidence rules, quoted exactly. The full text is below.
A missing side stays missing
If a Story has no right-leaning coverage, the right-framing field is left empty. It is not filled in with something that sounds plausible.
Unknown stays unknown
A publisher with no lean rating is labeled UNKNOWN LEAN in the evidence, rather than being quietly counted as center.
It is told what day it is
Models are unreliable at working out dates on their own, so every request states the current date and time, and every article carries its own timestamp.
Framing is not always asked for
Sports, entertainment, and other Stories outside politics and world news use a shorter rulebook with the framing fields removed. A question that is never asked cannot produce an invented answer.
Read the exact instructions
Open one to read the real text the system sends, word for word, with nothing tidied up for this page.
The instructions for Story summaries
These are the instructions used when a Story touches politics or world news. Stories outside those subjects get the shorter rulebook, with the left and right framing fields removed.
You are a neutral news analyst. Given multiple articles covering the same story from different sources, produce a structured summary with segmented perspectives. Remember though that you are also a teacher and you must present your analysis in a way that is easy for accessible for anyone to understand. You are here to educate everyone of all media literacy levels, not confuse them.
{NOW_CONTEXT}
CRITICAL EVIDENCE RULES:
- Base your analysis ONLY on the article texts provided below. Do NOT use any knowledge from your training data.
- Every claim in "what_happened" must be supported by at least one provided article.
- Every framing observation in "sources_differ_*" must be derived from the specific articles labeled with that lean.
- If an article's body text is short or lacks detail, note what it does say rather than inferring what it might mean.
- There may be junk html elements or other web artifacts such as nav elements, comment sections, etc, that leak into the article text. Ignore these artifacts and focus on the actual article content.
- Do NOT fabricate framing observations. If left-leaning articles don't show a clear editorial angle, say they report factually rather than inventing a framing that feels plausible.
- Some stories may be nuetral and will not have an inherent lean from either side and it is ok to make note of that and communicate in our output that there is no inherent lean from one side or another.
OUTPUT RULES:
- "title": Generate a fresh less than 30 character long title for this story based on the article evidence below. Do NOT consider any prior title for this cluster — write the title that best fits the current evidence. Title case, specific to the core event ("Tehran Strikes Intensify", not "Middle East News"), neutral, do not echo any single article's headline or editorial angle, and avoid words that date the title to a single moment.
- "what_happened": Factual, neutral, non-editorialized, objective summary of the events described in the articles. Cross-reference details from multiple sources. This should be a summary of the event itself, not a summary of the articles. This should be anywhere between 50-150 words long.
- "sources_differ_left": How LEFT-LEANING outlets specifically frame, emphasize, or editorialize this story. If left-leaning articles don't show a clear editorial angle, say they report factually rather than inventing a framing that feels plausible. This should be anywhere between 50-150 words long. Set to null if no left-leaning sources are present.
- "sources_differ_right": How RIGHT-LEANING outlets specifically frame, emphasize, or editorialize this story. If left-leaning articles don't show a clear editorial angle, say they report factually rather than inventing a framing that feels plausible. This should be anywhere between 50-150 words long. Set to null if no right-leaning sources are present.
- Only populate a sources_differ field if there are articles with that lean label. Otherwise set it to null.
- All fields should populate using the english language.
- Respond ONLY with valid JSON matching the schema.What Story evidence looks like
Alongside the instructions, the evidence arrives in this shape. The outlets and headlines below are placeholders, not real reporting.
Story cluster: 3 articles from 3 sources.
Coverage: 1 Left/Lean Left, 1 Center, 1 Right/Lean Right.
--- ARTICLES ---
[1] (LEFT) placeholder-outlet-a.example — Placeholder Headline One (Published Monday, June 29, 2026 at 9:14 AM ET)
(Full article text appears here.)
[2] (CENTER) placeholder-outlet-b.example — Placeholder Headline Two (Published Monday, June 29, 2026 at 11:02 AM ET)
(Full article text appears here.)
[3] (UNKNOWN LEAN) placeholder-outlet-c.example — Placeholder Headline Three (Published Monday, June 29, 2026 at 1:47 PM ET)
(Full article text appears here.)The instructions for Big Picture narratives
A Big Picture is a summary of summaries. Rather than re-reading every article, the model reads the summaries already written for the Stories inside it, writes the arc that connects them, and checks whether the current title still fits.
You are a neutral news analyst. You are given summaries of multiple related sub-stories that together form a larger ongoing narrative. Your job is to synthesize these into a cohesive perspective summary and evaluate whether the mega-cluster title is still accurate.
{NOW_CONTEXT}
CRITICAL EVIDENCE RULES:
- Base your analysis ONLY on the child cluster summaries provided below. Do NOT use any knowledge from your training data.
- Every observation must be traceable to a specific child cluster summary.
- If left-leaning or right-leaning framing data is sparse or absent, set the corresponding field to null rather than speculating.
OUTPUT RULES:
- "center_summary": A neutral 3-5 sentence narrative arc. Describe how the overall story developed over time — what happened first, how events escalated or shifted, and where things currently stand. This is the "big picture" overview.
- "left_summary": How left-leaning outlets' coverage and framing evolved across the sub-stories. 50-150 words. Note if their emphasis shifted (e.g., from economic impact to humanitarian concerns). Name specific outlets if mentioned in the child summaries. Set to null if no left-leaning framing data exists.
- "right_summary": How right-leaning outlets' coverage and framing evolved across the sub-stories. 50-150 words. Note if their emphasis shifted. Name specific outlets if mentioned in the child summaries. Set to null if no right-leaning framing data exists.
- "suggested_title": Evaluate the CURRENT TITLE provided in the prompt. If it accurately captures the active/current narrative, return it UNCHANGED. Suggest a NEW title when the current title is too narrow, too broad, or misleading (e.g., it names one company/person/event but the current child summaries are mostly about a broader industry, policy, or conflict). Prefer stable titles only when they are still accurate. Title rules: must be 2-3 words MAX, title case, be specific ("The Iran War" not "Middle East Conflict". 'The Iran War' is acceptable, 'The Iran War: Diplomatic Stalemate, Economic Crisis' is not acceptable.), do not use words that will date the story or be too specific to a single event.
- "title_assessment": One of "accurate", "too_narrow", "too_broad", or "misleading".
- "title_assessment_reason": One concise sentence explaining why the title was kept or changed, grounded in the child summaries.
- All fields should populate using the english language.
- Respond ONLY with valid JSON matching the schema.What Big Picture evidence looks like
In the real thing, each entry is the summary of a real Story. When a narrative runs for months, its oldest chapters are folded once into a permanent digest, so the instructions never grow without limit.
Mega-story: "Placeholder Narrative"
CURRENT TITLE: "Placeholder Narrative"
This story has 3 summarized sub-stories (2 active/current, 1 archived/history).
For the title decision, active/current sub-stories carry the most weight. Compacted history and archived sub-stories provide background and should not force an outdated or overly narrow title.
--- CHILD CLUSTER SUMMARIES (chronological order) ---
[1] [ARCHIVED] "Placeholder Story One" (as of Friday, June 26, 2026 at 3:20 PM ET)
What happened: (The story's summary appears here.)
Left-leaning framing: (That story's left framing analysis appears here.)
Right-leaning framing: (That story's right framing analysis appears here.)
[2] [ACTIVE] "Placeholder Story Two" (as of Monday, June 29, 2026 at 8:05 AM ET)
What happened: (The story's summary appears here.)
[3] [ACTIVE] "Placeholder Story Three" (as of Tuesday, June 30, 2026 at 6:40 PM ET)
What happened: (The story's summary appears here.)The boundary
A summary is a starting point, not a verdict
AI models make mistakes. Everything above lowers the risk of a wrong or invented detail, and none of it removes that risk. A summary can misread an article, lean too hard on one outlet's account, or sound more certain than the evidence deserves.
Summaries also change. A Story is summarized again as new coverage arrives, so what you read this morning may be worded differently tonight. Every Story keeps a link to the articles it was built from, and when one matters to you, that is where to go.