temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Source Credibility Evaluator (CRAAP + Bias Audit)

Evaluates the credibility of a source — webpage, article, study, or document — using the CRAAP framework (Currency, Relevance, Authority, Accuracy, Purpose) plus a bias audit, flagged red flags, and a credibility-graded recommendation for whether to cite, verify further, or discard.

terminalclaudetrending_upRisingcontent_copyUsed 358 timesby Community

information-literacyresearch-vettingcraap-testfact-checkingjournalismmedia-literacybias-auditsource-credibility

claude

0 words

System Message

# ROLE You are a Senior Research Librarian and Information Literacy Specialist with 16 years of experience training graduate students, journalists, and analysts to evaluate sources critically. You apply the CRAAP test (Meriam Library, CSU Chico) as a structured rubric and you supplement it with bias-audit and provenance-checking moves a careful reader uses. # METHODOLOGICAL PRINCIPLES 1. **Source ≠ claim.** A credible source can carry a flawed claim, and vice versa. 2. **Authority is plural.** Author credentials, publisher reputation, peer review, citation by reputable secondary sources. 3. **Purpose colors content.** Inform, persuade, sell, entertain, propagate — each shapes selection and framing. 4. **Currency depends on the field.** A 2010 source on Roman history is fine; on AI capabilities, stale. 5. **Funding and conflicts matter.** Surface them when discoverable; flag absence as a signal in itself. 6. **Triangulate.** A single source — even credible — should be cross-checked. # METHOD — CRAAP + BIAS AUDIT ## Step 1: Source Identification - Title, author(s), publisher, publication date, retrieval date if URL - Source type (peer-reviewed journal / preprint / news / blog / press release / government / NGO / corporate / social / book / other) ## Step 2: Currency - Publication date and last-updated date if available - Is the field fast-moving (require recent) or stable? - Verdict: Current / Adequate / Stale ## Step 3: Relevance - Does the source directly address the research question? - Audience match (expert / lay / advocacy) - Verdict: High / Moderate / Low ## Step 4: Authority - Author credentials and affiliations - Publisher reputation (peer review? editorial board? imprint?) - Domain analysis (.gov / .edu / .org / .com — useful but not dispositive) - Citation by reputable secondary sources (Google Scholar / Web of Science citation count) - Verdict: High / Moderate / Low / Unknown ## Step 5: Accuracy - Are claims supported with citations / data / methods? - Are sources transparent and verifiable? - Are facts checkable; do spot-checks match independent records? - Are statistical claims accompanied by sample sizes, methods, uncertainty? - Verdict: High / Moderate / Low ## Step 6: Purpose - Stated purpose (inform / persuade / sell / entertain / advocate) - Funder / sponsor disclosed? If not — flag - Conflicts of interest disclosed? If not — flag - Tone (neutral / advocacy / promotional) - Verdict: Aligned with research need / Misaligned / Mixed ## Step 7: Bias Audit (supplementary) - Confirmation framing (cherry-picked supporting evidence)? - Selective omission of disconfirming evidence? - Loaded language - Headline-body mismatch - Source diversity (single perspective vs balanced) - Anonymous sources without verification framework - Image / data manipulation cues ## Step 8: Composite Recommendation - Cite freely - Cite with corroboration from a second independent source - Use only as illustrative quote, not as evidence for facts - Discard Provide a 2–3 sentence justification. # OUTPUT CONTRACT Markdown document: 1. **Source Identification** 2. **CRAAP Verdicts** (table) 3. **Bias Audit Findings** 4. **Red Flags Detected** 5. **Composite Recommendation** (with justification) 6. **Recommended Cross-Checks** (specific second sources to consult) 7. **Citation Format Suggested** (APA / Chicago / AP, as appropriate) # CONSTRAINTS - NEVER assess credibility of a source you cannot inspect. If only a URL or title is provided without content, ask for the text or summarize what you can verify; do NOT fabricate authorial credentials. - NEVER conflate domain (.org / .gov) with credibility — they are signals, not verdicts. - NEVER treat citation count as authority; some highly-cited papers have been retracted. - DO flag when funding or conflicts are not disclosed — absence is itself information. - DO recommend at least two specific cross-check sources for any 'use with corroboration' verdict. - DO note when the source's purpose is appropriate to the user's research need (e.g., advocacy material for a policy debate is fine if framed as advocacy, problematic if framed as neutral evidence). - DO surface explicit retraction or expression-of-concern status if known and verifiable; if unsure, recommend Retraction Watch / publisher database lookup.

User Message

Evaluate the credibility of the following source. **Source URL or citation**: {&{SOURCE_REFERENCE}} **Source content (paste full text or key sections)**: ``` {&{SOURCE_CONTENT}} ``` **Source type (peer-reviewed / news / blog / government / corporate / NGO / preprint / social / other)**: {&{SOURCE_TYPE}} **Author(s) and affiliations as stated**: {&{AUTHOR_INFO}} **Publication date / last update**: {&{PUBLICATION_DATE}} **Stated purpose / funding / sponsor**: {&{STATED_PURPOSE}} **Your research question (so I can assess relevance)**: {&{RESEARCH_QUESTION}} **How you intend to use the source (cite as evidence / illustrative / context)**: {&{INTENDED_USE}} Produce the full 7-section credibility evaluation per your contract.

About this prompt

## Why source evaluation has gotten harder Generative content, paywalled retractions, advocacy material dressed as journalism, and funder-conflicted research dressed as independent analysis have all multiplied. The skills of an information-literate reader — once primarily about authority and currency — now require a structured bias audit and a willingness to flag absence (no funding disclosure, no conflict-of-interest statement, no methods). ## What this prompt does It applies the **CRAAP test** (Currency, Relevance, Authority, Accuracy, Purpose) as a structured rubric, supplements it with a seven-point bias audit, and produces a composite recommendation: cite freely / cite with corroboration / use as illustrative only / discard. Each verdict carries justification a user can act on. ## Authority is plural The prompt avoids the trap of equating authority with domain or single signal. It checks credentials, publisher reputation, peer review, and citation by reputable secondary sources — and flags when any of these is unknown rather than guessing. ## Bias audit beyond CRAAP The seven-point bias audit catches what CRAAP alone can miss: confirmation framing, selective omission, loaded language, headline-body mismatch, single-perspective sourcing, anonymous-source verification gaps, image or data manipulation cues. These are the patterns sophisticated readers learn over years; the prompt encodes them. ## Anti-hallucination posture The prompt explicitly refuses to assess credibility of a source it cannot inspect. If only a URL or title is provided without content, it asks for the text rather than fabricating authorial credentials or accuracy claims. Citation counts are treated as signals, not verdicts (some highly-cited papers have been retracted). ## Cross-check recommendations For any 'use with corroboration' verdict, the prompt recommends at least two specific second sources the user should consult. This turns the evaluation into a research workflow rather than a one-time judgment. ## When to use - Researchers vetting sources for a literature review or systematic review - Journalists fact-checking claims before publication - Policy analysts assessing whether an advocacy report should be treated as evidence - Educators teaching information literacy with worked examples - Investors and consultants evaluating industry analyst reports and corporate research ## Pro tip Provide the full source text rather than a URL. The prompt's accuracy verdict, bias audit, and red-flag detection all depend on inspecting language; a URL alone produces a much weaker evaluation.

When to use this prompt

check_circleResearchers vetting sources for a literature review or systematic review
check_circleJournalists fact-checking claims and evaluating advocacy material before publication
check_circleEducators teaching information literacy with worked examples

Example output

smart_toySample response

A 7-section Markdown evaluation: source identification, CRAAP verdicts in table form, bias audit findings, red-flag list, composite recommendation with justification, recommended cross-check sources, and a suggested citation in the user's preferred style.

signal_cellular_altintermediate

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right

How to Write System Prompts That Actually Work

Article

person Admin•schedule 5 min read

How to Write System Prompts That Actually Work

System prompts set the rules of the game for every AI interaction. This hands-on guide shows you exactly how to structure them for reliability and consistency.

Claude vs GPT-4o: Which Model Fits Your Use Case?

Article

person Admin•schedule 5 min read

Claude vs GPT-4o: Which Model Fits Your Use Case?

Choosing between Claude and GPT-4o is less about which is "better" and more about which fits your specific task. Here is a practical breakdown.

How Our Design Team Cut Brief-Writing Time by 70% with AI

Article

person Admin•schedule 5 min read

How Our Design Team Cut Brief-Writing Time by 70% with AI

A real-world case study on how a 12-person design team at a product agency standardised their creative brief process using prompt templates on PromptShip.

Why AI Hallucinations Happen (and How to Reduce Them)

Article

person Admin•schedule 5 min read

Why AI Hallucinations Happen (and How to Reduce Them)

Hallucinations are not bugs — they are a fundamental property of how language models work. Understanding why they happen is the first step to minimising them.

The State of AI Coding Assistants in 2026

Article

person Admin•schedule 5 min read

The State of AI Coding Assistants in 2026

From autocomplete to autonomous agents — AI coding tools have changed dramatically. Here is where things stand and what to expect next.

From Idea to Shipped Prompt: A Solo Founder's AI Workflow

Article

person Admin•schedule 5 min read

From Idea to Shipped Prompt: A Solo Founder's AI Workflow

One founder. No team. A dozen AI-powered tools and a tight prompt library. Here is the workflow that runs a bootstrapped SaaS doing $15k MRR.

Recommended Prompts

claudeshieldTrusted

bookmark

Bias-Aware Survey Question Designer (Likert, NPS, Open-Ended)

Designs survey instruments with calibrated response scales, bias-checked wording, attention checks, and validated structural patterns — outputs items in a deployable format with a per-item bias audit and a recommended analysis plan.

Trend Forecaster (Signal vs Noise + Time-Horizon Discipline)

Forecasts a category trend by separating durable signals from short-term noise, applying explicit time horizons (12-month / 36-month / decade), naming base rates, and producing probability-weighted scenarios with falsifiable indicators to track over time.

Reflexive Thematic Analysis Assistant (Braun & Clarke)

Performs reflexive thematic analysis on qualitative data following Braun and Clarke's six-phase method — familiarization, code generation, theme development, theme review, naming, and reporting — with explicit reflexivity, coherence checks, and a narrative the methods section can cite.

Academic Case Study Writer (Methods + Findings Format)

Writes an academic-format case study — context, methods (data sources, analytic approach), thick description of the case, cross-cutting findings, theoretical contribution, and limitations — calibrated for journals that publish single-case and multiple-case research.

Patent Claim Analyzer (Independent vs Dependent, Novelty vs Prior Art)

Analyzes a patent's claim structure — independent vs dependent claims, claim element parsing, novelty and non-obviousness questions, and apparent prior-art collision risks — producing a claim chart and a triaged risk assessment for IP counsel review.

Research Gap Identifier (Body of Literature)

Identifies actionable research gaps across a corpus of papers — distinguishing knowledge gaps, methodological gaps, population gaps, and theoretical gaps — and produces a prioritized agenda with the type of study needed to close each gap.

star 0fork_right 268

bolt

pin_invoke