30 Claude prompts that write annotated bibliography entries you can defend

<context> I have several sources for an annotated bibliography on [RESEARCH QUESTION] for [COURSE OR PROJECT]. I need thematic clusters, not an author-by-author list and not a fake literature map. </context> <inputs> [PASTE SOURCE] Paste each source as a block: citation fields plus abstract or my notes. Number them Source 1, Source 2, and so on. How many sources I pasted: [NUMBER] Citation style for the final list: [CITATION STYLE] </inputs> <task> 1. Propose 3 to 5 themes that are actually present in the pasted blocks. 2. Place each numbered source under one primary theme (and a secondary theme only if the paste supports it). 3. For each theme, write a 2-sentence cluster note: what these sources share, and one tension inside the cluster. 4. Flag any source that does not fit a theme instead of forcing it. </task> <constraints> - Use only the pasted sources. Do not add papers to fill a theme. - Theme names must come from claims in the paste, not from a generic textbook list. - If I pasted fewer than 3 sources, say clustering is premature. - Do not invent agreement that is not on the page. </constraints> <format> Theme name Sources in this theme (numbers + short titles from the paste) Cluster note Leftover sources that do not fit </format>

<context> I need my annotated bibliography to show a conversation, not a stack of summaries. Question: [RESEARCH QUESTION]. Project: [COURSE OR PROJECT]. I will only compare sources I pasted. </context> <inputs> [PASTE SOURCE] Paste 2 or more sources with the claim I think each makes. The specific point I want compared: [CLAIM OR VARIABLE] </inputs> <task> Build two clusters from the paste only: 1. Agreement: sources that actually say compatible things about [CLAIM OR VARIABLE], with a short quote or paraphrase from each paste. 2. Disagreement: sources that conflict on that same point, with the conflict stated in one sentence. Then write one annotation add-on I can append to each source: "Agrees with / Disagrees with Source N because..." </task> <constraints> - If two sources do not discuss the same point, say they are not comparable. - Do not invent a debate to make the bibliography look sophisticated. - Do not add a third scholar who is not in the paste. - Keep each add-on to 2 sentences. </constraints> <format> Agreement table (source, pasted evidence) Disagreement table (source, pasted evidence, conflict sentence) Per-source add-on lines Not comparable (if any) </format>

<context> I want my annotated bibliography on [RESEARCH QUESTION] grouped by method family (for example interview, survey, experiment, archival, theoretical) so I can see what kind of evidence I actually have. </context> <inputs> [PASTE SOURCE] Paste each source with any methods sentences I have. Method labels I am allowed to use: [YOUR LIST, e.g. qualitative interview, survey, experiment, theoretical, unknown] </inputs> <task> 1. Assign each pasted source to one method family from [YOUR LIST], or to "unknown" if the paste has no method. 2. Write a cluster note per family: what this method can show for [RESEARCH QUESTION], and what it cannot. 3. For each source, add one evaluation sentence about method fit that I can drop into the annotation. 4. State which method families are missing from my paste, without naming papers to fill them. </task> <constraints> - Do not guess a method that is not in the paste. - Do not invent a study that uses the missing method. - "Unknown" is a valid bucket. - Keep notes tied to [RESEARCH QUESTION]. </constraints> <format> Method family Sources assigned Cluster note (can / cannot show) Per-source evaluation sentence Missing families in this paste (no suggested citations) </format>

<context> My sources are currently in chronological order. For [COURSE OR PROJECT] on [RESEARCH QUESTION] I need to know whether a thematic annotated bibliography is more honest than a timeline. </context> <inputs> [PASTE SOURCE] Paste each source with year and a 1 to 2 sentence note on its claim. Current sort: chronological. Instructor preference if any: [THEMATIC / CHRONOLOGICAL / UNSPECIFIED] </inputs> <task> 1. Keep a dated list from the years in the paste only. 2. Propose a thematic regroup and show which sources move. 3. Write a 3-sentence rationale: when chronology helps this set, when theme helps, and which I should use for the submitted bibliography. 4. Draft a one-line headnote I can place above the list explaining the chosen order. </task> <constraints> - Do not invent years or "seminal" earlier works. - If dates are missing, flag those sources instead of dating them. - Do not add historical background sources. - The headnote must not claim a field history I did not paste. </constraints> <format> Dated list (from paste only) Thematic regroup Recommendation (theme or chronology) with rationale Headnote for the submitted list </format>

<context> I think my sources on [RESEARCH QUESTION] sit in different schools or frameworks. I need clusters I can defend from the paste, not a tour of theories Claude recalls. </context> <inputs> [PASTE SOURCE] Paste each source plus any named theory, framework, or key terms in the text. Candidate school names I am considering: [SCHOOL NAMES OR "NONE YET"] </inputs> <task> 1. For each source, quote or tightly paraphrase the language that signals a school or framework. 2. Cluster sources that share that language. 3. Reject any candidate name in [SCHOOL NAMES OR "NONE YET"] that the paste does not support. 4. Write a 2-sentence cluster annotation I can reuse under each source in that school. </task> <constraints> - Do not assign Foucault, Marx, Bourdieu, or any other name unless it appears in the paste or in my candidate list AND the paste supports it. - Do not invent a school to make the set look complete. - If the sources are atheoretical, say so. - Keep clusters smaller rather than forcing a fit. </constraints> <format> Cluster name (only if supported) Evidence language from each source Reusable 2-sentence cluster annotation Rejected school names and why </format>

<context> I need an argument map for my annotated bibliography on [RESEARCH QUESTION] so each entry can point to a claim, not just a topic. Project: [COURSE OR PROJECT]. </context> <inputs> [PASTE SOURCE] Paste each source with the sentences that look like claims. My working thesis: [WORKING THESIS] </inputs> <task> Build a claim map: 1. Extract 1 to 3 claims per source, using only wording supported by the paste. 2. Tag each claim as supports, complicates, or is orthogonal to [WORKING THESIS]. 3. Write a 1-sentence "use in paper" line per source that names the claim, not the topic. 4. List claims I wanted to find but that do not appear in the paste. </task> <constraints> - Do not upgrade a topic sentence into a stronger claim than the paste makes. - Do not add sources that would "complete" the map. - Orthogonal is allowed. Do not force relevance. - If I pasted only titles, refuse to extract claims. </constraints> <format> Source (from paste) Claims (1 to 3) Relation to thesis (supports / complicates / orthogonal) Use in paper (one sentence) Missing claims I still need to find myself </format>

<context> I want one annotated bibliography entry that shows how a single paper moves from evidence to claim. Question: [RESEARCH QUESTION]. Style: [CITATION STYLE]. </context> <inputs> [PASTE SOURCE] Paste abstract plus the results or argument paragraphs I copied. The claim I think the paper makes: [CLAIM I HEARD] </inputs> <task> 1. Restate the paper's claim only as far as the paste supports it. Correct me if [CLAIM I HEARD] is stronger than the paste. 2. List each piece of evidence in the paste and the claim it is asked to carry. 3. Mark evidence that is missing (mentioned but not shown) without inventing the missing numbers. 4. Write the annotation: citation, summary, evaluation of the evidence-to-claim link, use in paper. </task> <constraints> - Do not invent tables, quotes, or appendices. - If [CLAIM I HEARD] is not in the paste, say I misheard it. - Do not bring in other studies as evidence. - Keep the evaluation about this paper's link, not about the whole field. </constraints> <format> Corrected claim Evidence-to-claim table Missing evidence (named, not filled) Full annotation (citation, summary, evaluation, use in paper) </format>

<context> I need my annotated bibliography to hold the best case against [WORKING THESIS], using only sources I pasted. Project: [COURSE OR PROJECT]. Question: [RESEARCH QUESTION]. </context> <inputs> [PASTE SOURCE] Paste sources that might oppose me, including ones I hoped would help. Working thesis: [WORKING THESIS] </inputs> <task> 1. Identify which pasted sources actually oppose [WORKING THESIS], with evidence from the paste. 2. Steelman each opposing claim in 2 sentences. 3. Write an annotation add-on: how I will use this source as a counterargument in the paper, and what I still need before I rebut it. 4. If none of the sources oppose me, say my set is one-sided and do not invent an opponent. </task> <constraints> - Do not invent a critic or a famous rebuttal paper. - Do not weaken the opposing claim to make my thesis look safer. - If a source is mixed, split support vs opposition instead of picking a side. - No suggested reading list. </constraints> <format> Opposing sources (from paste only) Steelmanned claim Annotation add-on (use as counterargument) What I must still verify before I rebut One-sided set warning (if needed) </format>

<context> I need each annotated bibliography entry to name the warrant or assumption a source relies on, so I do not import hidden premises into [RESEARCH QUESTION]. Project: [COURSE OR PROJECT]. </context> <inputs> [PASTE SOURCE] Paste the argumentative paragraphs, not just the title. The claim I plan to borrow: [CLAIM I WANT TO USE] </inputs> <task> For each pasted source: 1. Name the stated claim I may borrow. 2. Name 1 to 3 assumptions required for that claim to hold, grounded in the paste (definition, population, time, measure, value). 3. Say whether those assumptions fit my project on [RESEARCH QUESTION]. 4. Write a 2-sentence evaluation I can paste under the citation. </task> <constraints> - Do not invent a theoretical lineage. - If the paste is too short to see a warrant, say so and stop. - Do not add sources that "share" the assumption. - Keep assumption language plain. </constraints> <format> Source Borrowable claim Assumptions (1 to 3) Fit to my project (yes / no / only if) Evaluation sentences for the annotation </format>

<context> I have a working thesis and a stack of sources. I need to see which parts of the thesis each source can carry before I write the annotated bibliography for [COURSE OR PROJECT]. </context> <inputs> [PASTE SOURCE] Working thesis broken into parts: [THESIS PART A], [THESIS PART B], [THESIS PART C] Paste each source with notes. </inputs> <task> 1. Make a map: thesis part → sources that can support it, with a phrase from the paste as evidence. 2. Mark thesis parts with zero supporting sources. 3. Mark sources that do not support any part (candidates to drop). 4. Draft a "use in paper" sentence for each kept source that names the thesis part. </task> <constraints> - Do not invent a source for an empty thesis part. - Do not stretch a source onto a part it does not address. - Drop recommendations must be based on the paste, not on prestige. - No extra citations. </constraints> <format> Thesis part Supporting sources + pasted evidence phrase Empty parts (find sources yourself) Sources to drop Use in paper sentences </format>

<context> I need a gap note for my annotated bibliography on [RESEARCH QUESTION]. A gap is something my pasted sources do not cover. It is not a list of papers Claude thinks exist. </context> <inputs> [PASTE SOURCE] Paste every source I currently have. Scope I promised in the assignment: [SCOPE STATEMENT] </inputs> <task> 1. State what the pasted set actually covers (topics, methods, populations, dates). 2. List coverage gaps relative to [SCOPE STATEMENT] that are visible because the paste is silent. 3. For each gap, write a 2-sentence bibliography note I can add as a memo to myself (not as a fake entry). 4. Rank gaps by how much they threaten the paper if I submit now. </task> <constraints> - Do not name authors, titles, or years to fill a gap. - Do not say "the literature has shown" unless that sentence is in my paste. - If the set is too small to diagnose a field-level gap, say the gap is in my reading, not in the field. - No suggested citation list. </constraints> <format> What this set covers Gaps (ranked) Memo notes (not bibliography entries) What I will search for next (queries, not titles) </format>

<context> I suspect my sources on [RESEARCH QUESTION] over-represent one population or place. I need that written into the annotated bibliography as a limit, not as a slur on the authors. </context> <inputs> [PASTE SOURCE] Paste each source with any stated sample, country, years, or setting. The population or context my paper is about: [MY POPULATION OR SETTING] </inputs> <task> 1. Table each source's stated population, place, and years from the paste (or "not stated"). 2. Compare that table to [MY POPULATION OR SETTING]. 3. Write one shared "limit" paragraph I can reuse, plus a one-line limit for each source annotation. 4. Name the gap in plain language. Do not fill it with a study I did not paste. </task> <constraints> - Do not guess a sample from the journal name or author affiliation. - Do not invent a Global South study, a recent replication, or a local dataset. - "Not stated" is a valid cell. - Keep the tone factual. </constraints> <format> Population table Mismatch with my setting Shared limit paragraph Per-source one-line limit Search terms I can use (no titles) </format>

<context> My annotated bibliography on [RESEARCH QUESTION] may be method-heavy in one direction. I need a gap note about methods I do not have, without Claude inventing a paper that uses them. </context> <inputs> [PASTE SOURCE] Paste sources with any methods I can see. Methods my assignment or thesis needs: [NEEDED METHODS] </inputs> <task> 1. Count which needed methods appear in the paste and which do not. 2. Write a gap note that says what I cannot claim yet because a method is missing. 3. Add a sentence I can place at the end of the bibliography: the set's method limit. 4. Give search filters (method keywords) I can take to a database. </task> <constraints> - Do not invent a mixed-methods paper, an RCT, or an ethnography to balance the set. - Do not claim the field lacks a method. Claim only that my paste lacks it. - If methods are unstated, count them as unknown, not as a family. - No author names outside the paste. </constraints> <format> Methods present Methods missing from this paste What I cannot claim yet Closing sentence for the bibliography Database filters (keywords only) </format>

<context> Two or more of my sources on [RESEARCH QUESTION] seem to disagree, and I do not want the annotated bibliography to hide that. I also do not want Claude to "resolve" the fight with a paper I do not have. </context> <inputs> [PASTE SOURCE] Paste the conflicting passages. The point of conflict as I see it: [CONFLICT POINT] </inputs> <task> 1. Restate each side using only the pasted passages. 2. List possible reasons for the conflict that are visible in the paste (different measure, years, sample, definition). Label guesses as guesses. 3. Write a shared annotation paragraph that reports the contradiction and refuses to pick a winner without more evidence. 4. State what I would need to verify (a number, a definition, a third source I must find myself). </task> <constraints> - Do not invent a meta-analysis or a later paper that settled this. - Do not declare a winner to make my thesis tidier. - If the passages are not actually about the same point, say the contradiction is only apparent. - No suggested citations. </constraints> <format> Side A (from paste) Side B (from paste) Visible reasons vs guesses Shared contradiction paragraph What I must verify next </format>

<context> I am tempted to ask Claude for "the" papers that fill my gap on [RESEARCH QUESTION]. I will not do that. I need a gap memo I can attach to the annotated bibliography that tells me how to search, not what to cite. </context> <inputs> [PASTE SOURCE] Paste my current sources and the gap I think I have: [GAP STATEMENT] Databases I can use: [LIBRARY DATABASES] </inputs> <task> Write a gap memo that includes: 1. The gap in one sentence, grounded in what the paste does not cover. 2. Inclusion criteria for the next source I will accept. 3. 5 search strings for [LIBRARY DATABASES] using keywords from the paste and from [GAP STATEMENT]. 4. A rejection rule: I will not add any citation Claude names from memory. </task> <constraints> - Do not output a reading list, DOIs, or "you should look at Author (Year)". - If you (Claude) feel an urge to name a paper, replace it with a keyword. - Keep the memo under 250 words. - Make the rejection rule copy-pasteable into my notes. </constraints> <format> Gap (one sentence) Inclusion criteria Search strings Rejection rule </format>

<context> I must submit an annotated bibliography for [COURSE] with [NUMBER] sources on [RESEARCH QUESTION] in [CITATION STYLE]. I will paste only sources I have in hand. Claude must not pad the count. </context> <inputs> [PASTE SOURCE] Paste each required source as a numbered block. Rubric notes: [RUBRIC: length per annotation, must include evaluation, must include use, other] Working thesis: [WORKING THESIS] </inputs> <task> For each numbered source, write a complete entry: citation, summary, evaluation, use in paper tied to [WORKING THESIS] or to [RESEARCH QUESTION]. After the last real source, stop. If I pasted fewer than [NUMBER], list how many slots remain empty and do not fill them. </task> <constraints> - One entry per pasted source. Zero invented sources. - Match [RUBRIC: length per annotation, must include evaluation, must include use, other]. - Flag missing citation fields instead of guessing. - If two pastes are the same source, say so and keep one. </constraints> <format> Entry N Citation Summary Evaluation Use in paper Empty slots remaining (if any) </format>

<context> I need to learn the formula for an annotated bibliography in [COURSE] before I batch the rest. Style: [CITATION STYLE]. Question: [RESEARCH QUESTION]. Teach with my first real source, then give me a checklist for the others. </context> <inputs> [PASTE SOURCE] Paste source 1 only (citation fields plus abstract or notes). What my instructor called a good annotation: [INSTRUCTOR NOTE OR "UNKNOWN"] </inputs> <task> 1. Show the formula in 4 labeled parts: citation, summary, evaluation, use in paper. 2. Write source 1 with me, labeling each sentence as summary, evaluation, or use. 3. Point out one sentence I should rewrite in my own words to avoid close paraphrase of [PASTE SOURCE]. 4. Give a checklist I will apply to sources 2 through N myself. </task> <constraints> - Do not write entries for sources I did not paste. - Do not invent instructor rules. If [INSTRUCTOR NOTE OR "UNKNOWN"] is unknown, stick to the four-part formula. - Keep the teaching voice short. - Do not add "model" sources. </constraints> <format> Formula Worked entry for source 1 (sentences labeled) Rewrite this sentence (close-paraphrase risk) Checklist for the rest </format>

<context> My [COURSE] annotated bibliography requires [WORD COUNT] words per entry, plus [CITATION STYLE] citations. Question: [RESEARCH QUESTION]. I tend to write summaries that eat the whole word count and skip evaluation. </context> <inputs> [PASTE SOURCE] Paste one source. Required parts: [SUMMARY / EVALUATION / USE] with target words: [SPLIT, e.g. 50 / 50 / 30] </inputs> <task> Write one entry that hits [WORD COUNT] within 10 words, with the split I named. Then show a word count per part. If the paste cannot support an honest evaluation of that length, write a shorter evaluation and say I must read more, rather than padding with generic praise. </task> <constraints> - No filler about how "important" or "seminal" the work is unless the paste supports that. - No extra sources to bulk the note. - If I asked for more words than the paste can honestly fill, stop early and say why. - Keep use in paper specific to [RESEARCH QUESTION]. </constraints> <format> Citation Summary (word count) Evaluation (word count) Use in paper (word count) Total Honesty note if the paste is too thin </format>

<context> I have weekly reading notes for [COURSE] and I need to turn only the readings I actually did into annotated bibliography entries on [RESEARCH QUESTION] in [CITATION STYLE]. </context> <inputs> [PASTE SOURCE] Paste my weekly notes. Each note should name the assigned text. Weeks to include: [WEEK LIST] Readings I skipped: [SKIPPED TITLES OR "NONE"] </inputs> <task> 1. Make one entry per reading that appears in the notes and is not on the skipped list. 2. Pull summary points from my notes; do not upgrade them into findings I did not write. 3. Add evaluation and use in paper only when my notes support it. Otherwise mark "need to reread." 4. Do not create entries for skipped weeks. </task> <constraints> - My notes are the source of truth, not Claude's memory of a famous syllabus. - If a note is only "interesting," refuse to write a summary. - Do not add recommended extras from the rest of the field. - Keep skipped titles out of the bibliography. </constraints> <format> Week Entry or "skipped / too thin" Need-to-reread flags </format>

<context> I am building a group annotated bibliography for [COURSE] on [RESEARCH QUESTION] in [CITATION STYLE]. Each member pastes only sources they personally checked. Claude must not merge in remembered citations to look complete. </context> <inputs> [PASTE SOURCE] Paste each member's sources as labeled blocks: [MEMBER NAME] then the source. Agreed thesis: [WORKING THESIS] Duplicate policy: if two members paste the same source, keep one and credit both names in a note. </inputs> <task> 1. Deduplicate by matching titles and authors in the paste. 2. Write one annotation per unique source, using only that block. 3. Add a contributor line (who pasted it). 4. List conflicts: two members summarized the same source differently. Do not pick a winner. Quote both notes. 5. Refuse any source that has no member label. </task> <constraints> - Do not add sources "the group should also include." - Do not invent a member. - If a block is only a URL with no notes, mark it incomplete. - Keep [WORKING THESIS] in each use-in-paper line. </constraints> <format> Unique entries (citation, annotation, contributor) Duplicates merged Summary conflicts (both notes quoted) Incomplete blocks </format>

<context> I need an annotated bibliography entry in APA 7 for [COURSE OR PROJECT] on [RESEARCH QUESTION]. I will paste the source. Claude will not recall a reference from training data. </context> <inputs> [PASTE SOURCE] Paste author names as printed, year, title, venue, volume, issue, pages, DOI or URL if printed on the source. Annotation type required: [SUMMARY + EVALUATION + USE] </inputs> <task> 1. Build the APA 7 reference from supplied fields only. Use the hanging-indent-ready single paragraph form. 2. If a required APA field is missing, write "MISSING: ..." and do not guess. 3. Write the three-part annotation under the reference. 4. List 3 APA 7 checks I should run against the PDF or library record (author initials, DOI, capitalization). </task> <constraints> - Do not retrieve or invent a DOI, page range, or issue number. - Do not convert a magazine into a journal. - Do not add extra references. - Keep the annotation free of APA myths that are not in my paste. </constraints> <format> APA 7 reference (or MISSING fields) Summary Evaluation Use in paper Checks against the source record </format>

<context> I need an MLA 9 works-cited entry plus annotation for [COURSE OR PROJECT] on [RESEARCH QUESTION]. Core elements must come from the source I paste, not from Claude's memory of a famous edition. </context> <inputs> [PASTE SOURCE] Paste title page, author, container (journal, site, book), publisher or venue, date, location (pages or URL), and access date if a web source. Container I think this is: [JOURNAL ARTICLE / CHAPTER / WEB PAGE / OTHER] </inputs> <task> 1. Order MLA 9 core elements from the paste. Stop at the first missing element and name it. 2. Write the three-part annotation (summary, evaluation, use in paper). 3. If I mislabeled the container, correct the label using only evidence in the paste. 4. Give 3 MLA checks (italics vs quotes, container, location). </task> <constraints> - Do not invent a publisher city, a 9th edition workaround, or a page count. - Do not cite a database name unless I pasted it. - Do not add other works. - Keep URLs only if I pasted them. </constraints> <format> MLA 9 works-cited line (or missing element) Corrected container type Summary Evaluation Use in paper Checks </format>

<context> I need a Chicago notes and bibliography pair plus an annotation for [COURSE OR PROJECT] on [RESEARCH QUESTION]. I am using notes-bibliography, not author-date, unless I say otherwise. </context> <inputs> [PASTE SOURCE] Paste all bibliographic facts I have. Note number I will use: [NOTE NUMBER] Fact I might footnote from this source: [FACT OR QUOTE I COPIED] </inputs> <task> 1. Write a full bibliography entry and a first-note form from the paste only. 2. Write a shortened note form that does not invent a short title. 3. Write the three-part annotation. 4. If [FACT OR QUOTE I COPIED] is not in the paste, refuse the note body and say I must paste the quote. </task> <constraints> - Do not invent page numbers for the note. - Do not switch to author-date. - Do not add a second source "for comparison." - If edition or editor is missing, mark MISSING instead of assuming 1st edition. </constraints> <format> Bibliography entry First note Shortened note Annotation (summary, evaluation, use in paper) Page number still needed (if any) </format>

<context> I need a Harvard author-date reference and annotation for [COURSE OR PROJECT] on [RESEARCH QUESTION]. Harvard variants differ by department, so I will paste my department's rules if I have them. </context> <inputs> [PASTE SOURCE] Department Harvard rules (if any): [PASTE STYLE NOTES OR "USE GENERIC AUTHOR-DATE"] In-text mention I want: [WHAT I WANT TO SAY IN THE SENTENCE] </inputs> <task> 1. Build the reference list entry from [PASTE SOURCE] fields only, following [PASTE STYLE NOTES OR "USE GENERIC AUTHOR-DATE"]. 2. Offer one in-text citation form (Author, Year) that does not invent a page. 3. Write the three-part annotation. 4. List fields my department sheet asks for that are missing from the paste. </task> <constraints> - Do not invent a Harvard variant (comma vs colon, ampersand vs and) if my sheet is silent. Pick one, label it "generic," and tell me to match the sheet. - Do not add sources. - Do not fabricate a page for the in-text cite. - If year is missing, use the missing-year rule from my sheet, or write n.d. only if I approve after seeing that the paste has no date. </constraints> <format> Reference list entry In-text form Annotation Missing fields vs department sheet Variant label (department vs generic) </format>

<context> I already have an annotation I trust. I need the same facts restyled from [CURRENT STYLE] to [TARGET STYLE] for [COURSE OR PROJECT]. Facts must not grow during the conversion. </context> <inputs> [PASTE SOURCE] Paste the current citation and the annotation I already wrote or approved. Current style: [CURRENT STYLE] Target style: [TARGET STYLE] </inputs> <task> 1. List every bibliographic fact present in the paste (author, date, title, container, pages, DOI, URL, edition). 2. Rebuild the citation in [TARGET STYLE] using only that fact list. 3. Keep the annotation text, only fixing in-text citation markers to the target style. 4. Output a "not converted" list for facts the target style wants that I did not paste. </task> <constraints> - Do not look up or recall missing facts. - Do not change the source's claims while restyling. - Do not add a second edition, a translator, or a DOI to be helpful. - If the annotation cites a source that is not in the paste, flag that sentence. </constraints> <format> Fact inventory Restyled citation Annotation with target in-text markers Not converted (missing facts) Flagged sentences </format>