Prompt Library

Develop a practical theory of change for the program described in my material. Use its objective, population, activities, intended outcomes and evidence. Keep the pathway within that scope.

Build a logframe from my program narrative, results chain, work plan and reporting requirements. Do not presume a sector, result count or organization-wide strategic objective.

Draft a practical monitoring, evaluation, and learning plan for the supplied program. Align it with the actual results chain and decisions the team must make. If the program purpose or results are missing, ask for them before drafting. Treat any activities, indicators, targets, dates, owners, and budget figures you propose as options awaiting approval.

Build an evaluation matrix from my program logic, evaluation purpose, intended users and decisions, priority questions and available evidence. Do not presume a sector, impact design, validated sources or numeric measures for every question.

Develop a focused indicator set for my objectives and timeframe, using any supplied theory of change and confirmed data sources. Do not assume a sector, baseline, target or impact-level result.

Create a reference sheet for each supplied or clearly proposed indicator. If there is no indicator statement or result to measure, ask for one before finalizing the sheet. Distinguish an existing indicator from your suggested revision.

Draft a learning agenda that helps the supplied team make specific decisions. If the program purpose and at least one decision or uncertainty are missing, ask for them before drafting questions. Do not claim stakeholder priorities or agreement without their recorded input.

Design a data-use plan centered on the decisions the supplied program actually faces. If no decision, intended user, or data source is identified, ask for the missing essentials before describing a real data flow. Treat new systems and roles as proposals, not existing arrangements.

Map the people and groups involved in, affected by, or able to influence the supplied program's MEAL decisions. If there is no program context or stakeholder evidence, ask for it before assigning real groups or power ratings. Separate documented roles from proposed groups to verify.

Build an activity-based MEAL budget for the supplied program. If activities, period, currency, or a basis for rates are missing, ask for them before giving a priced total. Never invent a local rate or call an unsupported figure realistic.

Build a results framework from the supplied program design. If activities or intended results are missing, ask for them before asserting a results chain. Show proposed links and assumptions as hypotheses until the program team validates them.

Design an accessible feedback and complaint mechanism for the supplied population and service. If the population, operating context, or response capacity is unknown, ask for it before selecting channels or response deadlines. Do not imply an existing mechanism was tested or endorsed.

Draft a factual progress report for my program and reporting period from supplied plans, monitoring results, activity records, financial data and reporting instructions. Do not assume a recipient, achieved outcome or approved format.

Draft an evaluation report from completed evaluation evidence. If findings or the methods that produced them are absent, ask for the report or evidence before writing results. Never fill findings with examples or imply impact from a descriptive design.

Write a concise executive summary of the supplied document for the named audience and decision. If no source document or substantive findings are available, ask the user to paste them before summarizing. Do not create findings, quantitative results, challenges, or recommendations to fill a section.

Write a focused case study from the supplied evidence for the named audience. If neither a documented case experience nor an implementation record is supplied, ask for source material before drafting a real case study. Do not invent a person, event, quote, outcome, or consent.

Draft terms of reference for an evaluation that a prospective evaluator could scope and price. If the evaluation purpose, subject, or intended users are missing, ask for them before presenting a commissionable TOR. Treat dates, budget, deliverables, approvals, and standards not supplied as proposed or to be confirmed.

Synthesize lessons from the supplied implementation evidence. If no experience record or observation is provided, ask for it before claiming lessons were learned. Do not invent consultations, causal links, or actions already taken.

Draft a concise MEAL section for the supplied proposal. If objectives or the proposal call are absent, ask for them before claiming compliance; a provisional section may be labeled as such. Do not infer an obligation from a funder name.

Plan how to share the supplied M&E findings so each audience can understand and use them. If findings have not been provided, ask for their source before drafting factual messages. If release clearance is unknown, label messages as internal drafts and identify the approval needed before external sharing. Never invent a metric, audience preference, prior feedback, or approval.

Review the supplied logframe as written. If no logframe is present, ask the user to paste it before making a document-specific assessment. Identify row or result labels when citing issues. Treat unprovided theory of change and program context as limits, not defects proven in the logframe.

Review the supplied MEL plan against its purpose, program design and operating constraints. If the plan is absent, ask for it before giving plan-specific findings. Request only the missing objectives, resources or reporting period needed for the assessment; do not assume a sector, mobile data capture or collection schedule.

Review the supplied theory of change against its stated purpose and context. Ask for the diagram or narrative if neither is supplied before judging its causal logic. Use only supplied material for claims about the program; examples and hypothetical results are not documented evidence.

Review the supplied indicator statements and definitions. If no wording is provided, ask for it. Assess only what the wording and supporting material establish, and request essential measurement context rather than guessing.

Review the supplied evaluation design against its questions, purpose and intended use. If the design is missing, ask for it before giving document-specific judgments. Assess the components actually present and identify what an excerpt cannot establish.

Review the supplied data collection instrument using its research questions, intended respondents, administration mode and prior feedback. Ask for the instrument if absent; a title or filename is not enough to review.

Review the supplied proposal MEAL section against the program objectives and any results framework, funding call, agreement or template provided. Ask for the section if absent; limit findings and revisions to the material available.

Review the supplied report draft for its intended readers. Use any program objectives, reporting requirements, previous feedback and priority areas provided. Ask for the actual draft if absent; a filename is not report content.

Analyze the supplied survey dataset and codebook or summary tables. If neither data nor summaries are available, ask for them before calculating. Use only supplied variables and comparison sources.

Analyze the supplied interview transcripts against the research question. Ask for transcripts if absent, and clarify the question if the material does not establish it. A file description is not evidence of themes or quotes.

Develop or refine a codebook for the supplied qualitative material and research question. If excerpts are absent, offer a provisional framework and request the material before coding participant data.

Summarize the supplied M&E report for its intended readers, using its scope and reporting framework unless another is provided. If the report is absent, ask for its text. Bound an excerpt-only summary to that excerpt.

Extract findings from the supplied document, using a focus question only if one is provided. Ask for the document if absent; a title alone is not source evidence. Limit extraction to the excerpt when the full document is unavailable.

Compare the supplied baseline and endline data or summary tables. Ask for either missing round; until both arrive, give only a comparison plan, not results.

Explore patterns in the supplied dataset or datasets. Ask for records or summary tables when only names or descriptions are provided. Analyze one dataset on its own if no second dataset exists.

Analyze gender-related participation, access and outcomes in the supplied data. Ask for data when absent and clarify the program question if the material does not establish it. Use only measured categories and supported comparisons.

Design a household survey using the supplied purpose, population, geography, constructs, collection mode and data-protection needs. Ask for essential missing inputs before calling the instrument field-ready. If a starter is requested, label it a draft and list the decisions and pretests still needed. Do not assume a sector, age range, topic list, question count, sampling frame or anonymity.

Design a semi-structured interview guide for the stated purpose, participant group and topic. Ask for these essentials if missing before writing topic-specific questions.

Design a focus group discussion guide for the stated group, topic and purpose. Ask for these essentials if absent before writing topic-specific questions.

Design a field observation checklist for the described activity and decision. Ask what will be observed and where if the activity is missing.

Design a sampling strategy for the supplied study purpose and target population. Ask for the population, intended estimate or comparison, and available frame if missing.

Design a baseline study for the supplied program. Ask for missing objectives, target population, location and intended future comparison before giving a program-specific design.

Design a rapid assessment for the stated urgent decision and setting. Ask for the decision, affected population or location, and deadline if absent before fixing collection and reporting plans.

Design a midterm evaluation for the supplied program. Ask for objectives, implementation stage, main decisions and available monitoring evidence when missing.

Explain the M&E concept I name in clear language. If I have not named a concept, ask which one before giving a lesson. Use any audience, work context or example I provide; otherwise use an accessible explanation without assuming a program.

Compare the named evaluation approaches for the supplied decision and program context. If no options are named, ask which matter or offer a clearly labeled short list to choose from.

Recommend an evaluation methodology for the supplied program, question and intended use. Ask for missing population, available data or major constraints before making a firm recommendation. Do not assume an impact evaluation or reject a sound single-method design.

Propose indicators for the supplied program objectives and primary outcome. Ask for missing objectives, outcome level, population, timeframe and available data before selecting program-specific measures.

Create an analysis plan for the supplied data type and question. First ask for missing question, observation unit, variables or qualitative material, collection design and intended audience. Do not assume a baseline survey, repeated measurements or trends.

Guide a contribution analysis for the supplied outcome, program theory and intended audience. Ask for missing observed outcome data, theory or evidence for causal links before writing a contribution story. Assess credible contribution, not proof of attribution; the six steps are iterative.

Create an adaptive management plan for the supplied program, phase and outcomes. Ask for missing decision rights, measures, review cycle and constraints before fixing triggers or a calendar. Unknown context remains unknown.

Create an unfilled after-action review template for the stated activity and organization. Ask for missing activity, participants, purpose and available duration before tailoring the guide. Do not invent results, causes, attendees, assigned owners or completed follow-up.

Draft a pause-and-reflect facilitation guide for the supplied program, review period and decision point. Ask for missing participants, evidence and available duration before fixing agenda times.

Create a learning brief for the supplied question, audience and decision date. Ask for the question and actual source material before writing findings. With partial sources, synthesize only those available and identify the gaps.

Assess whether to continue, adapt or stop the supplied intervention. Ask for missing decision criteria, actual performance, costs and authority before making a firm recommendation; provide a clearly provisional assessment when evidence is partial.

Draft the requested evidence brief, learning note or practice paper for the supplied topic and audiences. Ask for missing format, source findings and publication/privacy constraints. Draft only supported claims, leaving unavailable figures, quotations and approval status open.

Draft a structured lesson entry from the supplied activity and experience. Ask for missing context, observed result and evidence before writing a factual lesson. A possible explanation remains a hypothesis unless supported.

Design a Most Significant Change story-collection and selection guide for the supplied program. Ask for change domains, period, intended use, selection structure and consent/sharing rules before calling it field-ready.

Design an evidence-to-practice learning event for the supplied organization. Ask for missing audience, evidence, intended use, format and duration before fixing an agenda; leave dates unknown if not supplied.

Create an evaluation utilization plan centered on the supplied users and decisions. Ask for evaluation status, findings or expected outputs, decision authority/dates and disclosure constraints. With pending findings, make a prospective plan rather than inventing results.

Develop a GESI analysis framework integrated with the supplied M&E system. Ask for the results framework, population, context, known inclusion concerns and safe collection constraints before drafting program-specific indicators. Potential risk is not documented harm; do not infer sensitive identities, violence or significant disparities from a template.

Design a disability-inclusive data collection instrument that integrates WG questions appropriately and plans accessible participation of persons with disabilities throughout the data collection process. Label the instrument and protocols as proposals pending local review and testing; a drafted tool does not establish meaningful participation, validation or approval.

Analyze how overlapping identity dimensions relate to advantage or disadvantage in my supplied monitoring or evaluation data. Confirm actual program, data source, variable definitions and outcome indicators; do not import the examples as facts. Ask for data if absent. With partial data, identify unavailable comparisons and continue only with supported ones.

Create an environmental and climate screening checklist for my actual program activities and sites. Ask for descriptions, scale, jurisdiction/agreement and known concerns before classifying risk or compliance if absent. Screen plausible risks without presenting them as existing impacts. This draft grants no environmental clearance.

Create climate-sensitive indicators and a risk-monitoring framework that fit my existing M&E system. Ask for location, activities, theory of change, locally evidenced hazards, current indicators, decision cadence and climate-data access before specifying local measures/seasons. Do not assume the examples describe my program.

Analyze my program activities and monitoring evidence for conflict sensitivity using CDA Collaborative's Do No Harm framework. Ask for current context evidence, activities, monitoring material, affected groups, safety protocols and decision authority before drawing a program-specific conflict map. Use dated supplied/verified dynamics, not group stereotypes or assumed motives.

Design a protection mainstreaming monitoring framework for the program I describe. Ask for the activities, people reached, setting, existing safeguarding and complaint procedures, and decision purpose if they are absent.

Design a value for money framework for the program I describe. Request the program's outputs and outcomes, costing perspective, budget or cost sources, and decision purpose when missing.

Analyze the supplied cost and result data for a defined set of alternatives. Ask for the alternatives, cost boundary and perspective, output or outcome measure, periods, and data before calculating.

Design a locally led M&E plan for the program and community partners I describe. Ask who holds current decision and data rights, what communities want from M&E, and what reporting obligations actually apply if unknown.

Design an accountability to affected people framework for the program I describe. Ask for the population, services, communication access, existing channels, response capacity, and safeguarding rules before specifying channels or deadlines.

Review the M&E plan or instrument I provide for foreseeable harm. If no document is supplied, ask for it before making item-specific judgments; request the collection setting, participants, method, and intended data use when needed.

Build a contribution analysis framework for the program and outcome I describe. Ask for the intervention strategy, outcome evidence, time period, theory of change, and credible alternatives if absent.

Design a Most Significant Change protocol for the program, decision purpose, and reporting period I describe. Ask who will tell stories, who will select them, and how results will be used if unknown.

Design an Outcome Mapping framework for the program and the actors it directly influences. Ask for intended actors, vision, program strategies, and current relationships when absent.

Design an Outcome Harvesting protocol for the program and use I describe. Ask for the harvest question, scope, available records, likely changed actors, and reporting audience if absent.

Assess whether a quasi-experimental evaluation can credibly answer the supplied outcome question. Ask for assignment rules, intervention and comparison populations, baseline and follow-up data, timing, and constraints before recommending a design.

Develop and, when source texts are supplied, apply a qualitative coding and thematic analysis framework for the stated evaluation questions. Ask for the material and question if neither is given.

Design a developmental evaluation only if the supplied initiative is adapting in uncertain conditions and decision makers can use timely feedback. Ask for the innovation, decision cadence, evaluator access, and current evidence if missing.

Design a participatory evaluation for the stated purpose and groups. Ask who can participate safely, what authority participants will have, and how findings will be used before promising co-evaluation.

Apply realist analysis to a program with a clear explanatory question and actual outcome, context, and mechanism evidence. If only a program description is supplied, develop provisional theories and an evidence plan, not findings.

Design a process tracing protocol for the specified case and observed or proposed outcome. Ask for the outcome evidence, case chronology, program actions, and competing explanations when missing.

Design a mixed-methods evaluation for the questions and resources I provide. Ask what each qualitative and quantitative strand must contribute and whether integration is feasible before prescribing both.

Assess whether a proposed evaluation can credibly answer its intended questions. Ask for the program design, decisions, existing data, evaluation users, and constraints before assigning a readiness rating.

Plan engagement for the specified evaluation. Ask for the purpose, people affected, intended users, timeframe, power or safety concerns, and existing decision rights if absent.

Plan dissemination and use of evaluation findings. Ask for the source report or findings, audience decisions, release status, privacy rules, and timing before drafting factual messages.

Design a monitoring dashboard for my actual users and decisions. Ask for indicator definitions, source schemas/access, refresh/reporting cadence and privacy limits before a build-ready specification. Source/platform names do not establish working integration.

Create a data quality audit protocol for my indicators and reporting system. Ask for definitions, site/reporting frame, registers/forms, aggregates, period and governing agreement before operational steps. Names alone support a locally validated draft, not checked data or scores.

Draft an evaluation inception report from my actual TOR, program evidence and commissioning instructions. Ask for substantive TOR/questions, intended users, scope, schedule and evidence when absent. A title or yes/no answer is not TOR content.

Create an enumerator training plan for my actual instrument and field exercise. Ask for questionnaire, sampling/consent rules, language versions, team size and training time before a field-ready agenda/manual. A tool name does not establish skips; exercises, referrals and competency thresholds require local testing/approval.

Create a supervision checklist and data-quality spot-check protocol for my approved exercise. Ask for instrument, sampling/consent, back-check permission, configured platform, staffing and safety procedures before operational use. Local policy governs thresholds and personnel action; this draft proves no field readiness.

Draft a commissionable scope of work for my M&E assignment. Ask for purpose, background, scope, actual TOR/policy/procurement rules, budget and decision authority; policy/funder names alone establish no requirements. Mark unconfirmed dates, weights, payments, roles and deliverables proposed.

Create a technical scoring matrix from my full TOR/SOW, published solicitation/procurement rules and panel roles. A TOR title is insufficient. Do not evaluate firms or enter final scores without proposals and authorized reviewers; samples remain illustrative.

Draft a data management plan from my actual inventory, systems, team and requirements. Ask for tools, architecture, access roles, flows, retention, jurisdiction and agreement clauses before a deployable plan. Names are not configurations/rules; a draft establishes neither implementation nor compliance.

Create an audience-specific storytelling brief from my actual report/findings. Ask for source, verified findings, audience and decision first; a title is no evidence. Do not invent figures, motives, comparisons or quotes. Private language requires permission and safe disclosure.

Draft an outline to present my actual M&E findings. Ask for report/findings text, audience, time and intended decisions before slide messages; a title is not evidence. Leave unavailable figures, quotes, dates and approvals open. This is an outline, not a finished deck.

Draft infographic content from actual report/tables for my audience and channel. Ask for verified values with periods/denominators, brand constraints and quote permissions. Titles or claims lists cannot verify figures; never invent URL, contact, logo, quote or clearance.

Draft a two-page plain-language stakeholder brief from my actual evaluation/monitoring source. Ask for report/excerpts, audience, recommendations, distribution status and approved contact. A title is not evidence; leave unknown budget, partners, management response and contacts blank/flagged. Release remains pending until confirmed.

Recommend chart types for my variables/comparisons, audience, medium and tools. Ask for schema/definitions and intended relationships if absent; a dataset name is insufficient. Descriptions support chart selection, not unseen values/distributions/trends. Continue with classifiable variables and state missing details for others.

Draft a self-contained one-page program results snapshot for the stated period. Ask for actual indicator records, definitions, baseline/current/target periods, denominators and documented achievements. System names are not data; never invent contacts, logos, budget or clearance.

Draft a community feedback report from actual findings and verified response actions. Ask for source results, community language/access needs, safe channels and review permissions before public-ready copy. A survey title is not findings; do not promise confidentiality, no retaliation, translation accuracy or completed action without local confirmation.

Synthesize actual sources into an evidence brief for my question, scope, audience and decision. Request full text, accessible links, substantive extracts or abstracts; titles support only retrieval/screening, not ratings/findings. Distinguish rapid/purposive from systematic review.

Create an organizational M&E capacity assessment tool for the stated purpose, roles and safe participation arrangements. Ask for actual structure, jobs, documents and systems before tailoring. A maturity estimate is not a finding; leave scores and benchmarks blank until evidence is collected.

Develop an M&E capacity strengthening plan from my actual assessment findings, priority decisions, roles, budget, period and resources. Request missing evidence; an assessment title is not a gap analysis. Objectives, maturity targets, costs and owner commitments remain proposed until agreed.

Design a practical M&E curriculum for my actual audience roles, starting skills, available duration, delivery mode and learning goals. Ask for sector examples/materials if needed. Without essentials give a modular outline for customization, not a ready-to-run or already piloted/approved course.

Design an M&E mentoring/coaching framework for actual participant roles, goals, mentor availability, period, resources and confidentiality policy. Ask for missing essentials. Matching, commitments, ratings and governance are proposals, not participant consent or readiness.

Create a supportive partner M&E capacity assessment and strengthening plan from actual partner documents, staff/system evidence, agreements and support capacity. No rating/tiering from names or titles. Score bands, monitoring cadence, costs and reporting remain proposed until approved.

Draft a facilitation guide for my M&E session topic, actual time, audience, modality, goals and references. Ask for essentials before a ready-to-run script; otherwise label a modular example for adaptation. Do not pass fictional cases or unverified links off as local approved material.

Create an M&E competency framework for actual roles, job descriptions, HR constraints, core framework and staff consultation. Ask for missing essentials before setting level expectations. Pathways and criteria require HR/staff validation; years in post do not prove skill.

Design a practitioner-led M&E community of practice from actual member demand, staff time, platform/accessibility and budget. Request unknown essentials before commitments; sponsor, volunteer, funds and engagement remain proposals. Use relevant Wenger-Trayner/ALNAP models, not proof of local demand.

Design a digital form structure for my approved instrument/indicators, response options, sampling/consent rules, languages and chosen XLSForm platform. Request missing source text before deployable logic; platform or indicator names are not instruments. Expressions remain untested until validation and pilot in that tool.

Draft a mobile data collection protocol for my actual devices, form versions, connectivity, server/tenancy/security and support roles. Ask for missing configuration/access policy before field-ready steps. Names alone do not prove backup, encryption, remote wipe or sync; mark procedures untested until device/server validation.

Plan monitoring under my actual access constraints. Ask for indicator definitions/records, phone-list consent basis, imagery permissions, partner capacity and security conditions. Choose methods and callback rates for risk, design, consent and feasibility; shared biases can make nominally different sources nonindependent.

Design a monitoring system whose speed serves actual decisions. Ask for real feed/schema, latency, definitions/variation, approved thresholds, responders/capacity and channels. Source names are not live integrations; triggers/timing remain proposed until tested, funded and accepted. Avoid unreliable or sensitive alerts.

Draft an ethics-review submission package for my actual study and review body. Ask for protocol/instruments, consent/referral procedures, review instructions, qualifications and applicable law. Missing elements require responsible ethics/legal review; a draft implies no approval, compliance, referral readiness or certification.

Draft understandable consent materials for my actual activity, participant age/capacity, methods, duration, topics, risks/benefits, data practice, language and local review pathway. Ask for missing procedures, confidentiality limits, retention and contacts before usable text. Mark adaptation, translation and approval pending; promise no protection or withdrawal beyond actual procedure.

Draft a DPIA for actual processing purpose, data subjects/categories, lifecycle, vendors/partners, access/retention, jurisdiction and privacy governance. Request missing evidence; a technology name proves neither encryption nor hosting. Do not assert legal basis, DPIA duty, compliance or permission to proceed without responsible privacy/legal review.

Review my actual M&E plan, instruments, consent, data procedures, population/setting and governing policies for ethical gaps. Ask for missing essentials before rating compliance; titles or summaries cannot verify safeguards. Supporting records may inform judgments but do not make an undocumented plan section complete. Plans/policies alone prove neither implementation, legal compliance nor approval.

Design an M&E framework for actual health activities, results, geography/population, period and reporting needs. Ask for program logic, authorized HMIS dictionary/access, national standards/denominators and agreement text. DHIS2 names are not feeds or interoperability specifications; leave unverified targets/integration open.

Design a household health survey for actual objectives, respondent groups, geography/frame, indicators, source modules/languages and ethics pathway. Ask for missing essentials; named DHS/MICS surveys are not questionnaire text. Never invent validated wording, precision, anthropometry capacity or approval. Flag adaptation/pilot needs.

Analyze my actual facility-by-period extract, indicator dictionary, reporting expectations, population denominator sources, target definitions and service changes. Request missing evidence; a system name/screenshot is not data. Without data give an analysis plan/request, not trends/ranks/coverage. Mark unsupported partial figures not calculable and continue supported analyses.

Design an education M&E framework from actual activities, grades, school/learner frame, assessments/curriculum, period, EMIS schema/access and relevant standards. Ask for missing inputs. Names like EGRA/EMIS are not datasets or validated benchmarks; integration, cut scores and SDG mapping remain unverified until supported.

Design a literacy, numeracy or combined assessment blueprint for the actual learners, curriculum, grades and languages. Request the authoritative source form/items/version, intended inference, assessor capacity and locally validated cut-score evidence before field-ready specifications; an EGRA/EGMA name is not an item bank or benchmark file.

Design an M&E framework for the actual WASH activities, service area, population, management roles and decisions. Request applicable JMP module/source data, national standards, laboratory method, denominators and agreement sections before classifying service or compliance. Proposed monitoring is not observed service status or completed testing.

Design a household WASH survey for the actual purpose, population and service setting. Request the real frame, authoritative JMP module/version, observation protocol, languages and ethics safeguards before a field-ready tool. Do not presume camps, displacement or sensitive-status questions.

Design an M&E framework from the actual food-security/livelihoods objectives, activities, population, phases and modality. Request source instruments/editions, baseline or collection plans, frame and applicable agreement sections. Do not assume transfers, livelihood activities or an emergency-to-recovery sequence.

Analyze actual household and market data for the stated places/periods. Request records or sufficient tables, instrument versions, missingness, design/weights, prices/dates and any official IPC publication. Without data, give a calculation/quality-check plan rather than findings; with partial data, continue supported analysis and mark unavailable figures explicitly.

Design monitoring only for actual protection services and participant risks. Request the service model, indicator definitions, safeguarding/referral procedures, current access/disclosure rules, population context and applicable requirements. A CPIMS/GBVIMS/Primero name is not deployment evidence, case access or permission.

Help choose an evaluation approach for the supplied program and decision. Ask for missing purpose/questions, actual design, outcomes/activities, existing data and comparison options, users, budget, time and ethical/access constraints before a firm recommendation.

Review the supplied program data through completeness, accuracy, consistency, timeliness and validity. These are practical lenses, not one universal official framework; name a standard and edition only if supplied and applicable.

Extract a source-linked inventory of M&E requirements from the solicitation materials actually supplied. Ask for the full text, accessible annexes, amendments, referenced policy sections and due date before claiming complete coverage. A donor name, title or inaccessible link is not a requirement.

Assess whether the supplied donor opportunity's M&E requirements fit the organization's evidenced capacity, operating context and proposed scale. A No-Go is valid when supported; this is advice to the authorized capture decision-maker.

Assess credible M&E positioning for the supplied opportunity using the actual RFP/context, stated differentiators and accessible, dated public evidence. Ask for missing solicitation or competitor evidence before profiling named organizations. Without named or evidenced competitors, assess supported sector patterns rather than inventing bidders.

Turn the supplied M&E obligations, organizational capacity and program scale into a costable capture plan for budget and partner discussions. Request the actual solicitation, budget/duration/geography, available team, hiring rules, partner capacity and unit-cost basis. A summary cannot establish compliance with unseen sections.

Draft an M&E staffing plan sized to the actual MEL workload and program phases. Ask for the tasks, reporting frequency, collection rounds, availability, location, HR policy, rates, duration and budget basis. A program title or total budget alone cannot justify full-time positions or a staffing cost.

Write a proposal-ready M&E budget narrative linking each major cost to an actual MEL activity. Request the schedule, quantities, rates, currency/period, staffing plan, program total and applicable cost-rule text. A plan title or total alone cannot establish line-item costs.

Draft an indicator-by-indicator data collection plan for the actual MEL plan. Ask for indicator definitions, population/frame, intended estimands, instruments, reporting schedule, program duration, team, access and ethical constraints before fixing operational details. A named platform or instrument is not field-tested content.

Turn the actual supplied program narrative into a traceable theory of change. Ask for the narrative and supporting evidence before asserting a pathway or rating evidence strength; a title is not its contents. Use the stated long-term outcome and governing framework only where actually supplied.

Operationalize the supplied finalized theory of change into a logframe. Request the actual ToC, template, indicator definitions, baselines, data access and budget/timing constraints before calling it submission-ready. A title or template name is not its content.

Build a realistic drafting workflow within the actual remaining hours and available team. Ask for source documents, required deliverables, exact deadline, team/reviewer availability and usable prior materials. Do not assume 48 hours merely from the title or treat a summary as complete requirement evidence.

Stress-test the actual draft M&E section for supported logic, feasibility, compliance and credibility problems before submission. Ask for the draft and governing requirement text; a donor or document name is not compliance evidence. Be direct and specific, without inventing reviewer views or scoring rules.

Stress-test the actual M&E budget against MEL workload and documented cost rules. Request the spreadsheet/table with formulas or quantities, schedule, role LOE, comparable program total and binding rule text. A budget title, percentage or donor name cannot establish adequacy.

Capture specific lessons from the supplied project or milestone and connect them to future decisions.

Design an after-action review for the supplied completed activity, with practical facilitation and unconfirmed follow-up commitments.

Design a prioritized learning agenda for the actual program duration and geography, organized around decisions.

Design a reflection workshop around the team’s recent evidence and next decisions, fitted to the available time.

Synthesize evidence-based patterns and contrasts across the supplied portfolio without erasing context or overstating confidence.

Design a lightweight knowledge-management strategy around actual workflows, users, permissions and staff capacity.

Connect routine program monitoring to an operational learning loop with meaningful signals, decision rights and follow-up.

Distill policy-relevant insights from the supplied evidence for the stated topic, geography and policy decision.

Assess whether documented prior lessons changed current program design and practice, using observable uptake evidence.

Design a forward-looking pause-and-reflect session for the actual team, evidence and upcoming implementation decisions.

Adapt verified program lessons for donor, peer and internal readers without changing their evidence or certainty.

Document learning from program underperformance honestly, without blame and with explanations bounded by evidence.

Review my evaluation report draft against the supplied TOR, evaluation questions and relevant criteria. Use OECD DAC criteria only where appropriate to the evaluation purpose.

Review the actual MEL plan draft for its purpose and stage. Request missing text and relevant appendices, indicator sheets, instruments, role commitments, budget and submission criteria; filenames do not establish content.

Review the actual donor progress report for its reporting period. Request substantive text, applicable agreement/template sections and underlying indicator records; a report or funder name supplies none of them.

Review the actual questionnaire for its population and intended indicators. Request wording, response codes, routes, consent text and language versions, plus relevant definitions and ethics/privacy protocols. A tool name alone cannot support an item review.

Review the actual logframe with substantive ToC content and applicable agreement/template sections. Ask for missing text rather than inferring alignment from document or funder names.

Review the actual evaluation ToR before issue to bidders. Request substantive program background, procurement/evaluation-rule sections, scope, budget and timing where needed; titles are not source content.

Review the supplied indicator set and definitions with actual results, sources, schedules, targets/resources and governing requirements. Ask for missing source content; names alone do not establish definitions or obligations.

Review the actual inception report with its ToR, agreed scope changes, evaluation matrix, instruments, sample rationale and ethics/access records. Names or approval assertions are not those contents; request missing evidence and assess available parts.

Review major findings against their actual evidence. Request the findings text, methods/limitations and linked tables, excerpts or records; a citation or dataset title alone does not verify its contents.

Review the actual recommendations with linked findings/conclusions, decision users and known mandate, budget and timing constraints. Ask for missing text; a report title cannot establish actionability or feasibility.

Design an evaluation matrix showing how the actual priority questions will be answered and judged.

Design a sampling plan with defensible selection, size or qualitative scope, field procedures and inference limits.

Design a monitoring system that connects actual indicators and capture processes to useful, authorized decisions.

Describe the actual or proposed data flow with clear actors, source records, transfers, controls and use points.

Design a results framework for the supplied program scope, distinguishing intended changes from achieved results.

Review the actual triangulation section with its evaluation questions, method/source mapping, analytic approach and quality procedures. Ask for missing content; a triangulation claim or design label is not the process.

Review the actual sample rationale against study questions, population/frame or case-selection description, planned analyses, calculations and resource constraints. A sample total or study title is not justification; request missing evidence.

Review the actual limitations section alongside methods, findings and available response, fieldwork or cleaning evidence. With only the section, assess clarity/coherence and leave external completeness Not Assessable; request records needed for that comparison.

Review the actual analysis plan with questions, design/sampling details, variable definitions or safe qualitative descriptions and constraints. Ask for missing content; software or dataset names alone are not an implementable plan.

Review the supplied integration section with questions, design, sampling and analysis plans for both strands. Request missing content rather than inferring procedures from a mixed-methods label.

Review the actual engagement section with safe mapping summaries, decision scope/stage, feedback processes and implementation records where available. Ask for missing content; a plan title cannot establish participation or influence.

Review the actual operational sampling frame, separately from sample-size justification. Request its description/schema or safe aggregates, custodian/version, eligibility rules, target population and coverage/cleaning evidence. Registry names alone do not establish fit; do not request identifiable records where metadata suffice.

Draft an approach-appropriate qualitative coding frame, using provisional notes and reflexive memos when a fixed codebook does not fit.

Draft an adaptive-management plan linking actual monitoring triggers to authorized decisions and documented follow-up.

Design an instrument pilot that can reveal priority problems and support a later authorized readiness decision.

Review the actual indicator wording and definitions with intended results, measurement schedules and associated baseline, target and resource assumptions. Request missing content and assess each indicator/criterion only as far as evidence permits.

Review the actual indicator/definition table with baseline source or calculation notes, target/milestone periods and rationale, scope and resources. Dataset names alone do not verify values; request missing supporting content.

Review the actual indicator means of verification with definitions, source details and safe contextual evidence on access, resources, geography, security and quality. Request substantive records when needed; distinguish missing indicator documentation from information in separately cited supporting material.

Review only the supplied indicator statement's language. Ask for wording if absent and use a linked definition solely to clarify the unit, denominator or reference point, not to expand into a source, target, disaggregation or frequency audit.

Review the composite definition against its stated construct. Check theoretically grounded component coverage, named sources and redundancy; explicit equal/differential/hierarchical weights and their rationale/sensitivity; reproducible aggregation with normalization/scaling/transformation; missing-component rules and minimum completeness; and the substantive meaning of high/low scores. Transparent equal weights can be defensible; PCA is not automatically better.

Review each indicator's declared output, outcome or impact level against the actual results chain. Outputs are direct products/services; outcomes are changes in behavior, capacity, condition or knowledge; impact follows the broader/longer-term change defined by the supplied chain. Counts reached or trained do not by themselves measure changed behavior.

Review the supplied indicators against the actual applicable agreement, authoritative definitions and reporting template, recording titles, versions and sections. Request missing source text before judging compliance or equivalence. A funder name or current handbook does not establish applicability to every award.

Review the stated proxy and its justification. Check the specific barrier to direct measurement; the causal/correlational link defended by actual evidence or program logic; contextual/population/time/intervention limits; feasible and decision-useful direct validation, second measure or sensitivity check; and advance disclosure of measurement error, confounding, alternatives and attribution limits. Do not require direct measurement when unsafe or infeasible, or assume universal validity.

Review the document's disaggregation pipeline for its purpose and stage. Check necessary, safe and authorized identity variables; relevant indicator coverage; consistent categories/instrument capture; analysis; accessible protected reporting; and whether supported patterns inform findings, decisions, recommendations and next steps. Do not collect sensitive identities just to complete a checklist.

Review the summary as a standalone product for its intended decision-maker. Check whether program/evaluation purpose, scope, period and main conclusions are understandable without the full report; important supported findings are prominent; methods, findings, conclusions and recommendations receive proportionate coverage; jargon/acronyms and bureaucratic prose are explained; length fits audience/channel; and voice remains balanced about limits and underperformance.

Review each major finding and the evidence behind it. Check traceability to specific values/sources, safe quote or observation identifiers, or document pages; visible corroboration and contradictions; useful qual-quant integration where both exist; calibrated inference; explicit treatment of contrary evidence; and relevant safe subgroup patterns supported by the method.

Review each recommendation for an operational action and object; owner/decision authority or a linked management-response route; timing or priority; feasibility under actual resources/mandate/context; traceability to findings or another documented basis; and coverage without duplication. Flag vague aspirations, unsupported actions and important findings left unaddressed.

Review whether a competent peer can understand and critique the documented study and reproduce supported analytic steps, within confidentiality/access limits. Check named design and question-linked rationale; instrument purpose, format, length and respondents; frame, size, selection and applicable response rates; training/pilot/period/mode; raw-data-to-finding analysis for each data type; relevant QA/consent/privacy/referral safeguards; and honest implications of limitations. Do not require every method or promise anonymity from identifier removal.

Review whether concrete evidence-backed findings drive operational recommendations. Check finding specificity, evidence strength and relevant subgroup patterns; traceable recommendation links to findings or another documented decision basis; actionable what/who/when matched to authority; priority, feasibility, tradeoffs, sequencing/dependencies; and recorded follow-up/revision routes.

Review whether the needs assessment supports the intended program scoping and targeting decision. Check precise problem, population, geography, inclusion/exclusion and period; severity evidence appropriate to the decision; method/selection/source quality and limits; relevant needs, defensible magnitude or bounded qualitative description, safe subgroup patterns and prioritization; affected-population voice and GESI inclusion; and specific, feasible, traceable recommendations.

Review the supplied text for its intended non-specialist audience and use. Check readable vocabulary/sentences without a universal word limit; defined jargon/acronyms and helpful active voice; navigable headings and orientation where needed; useful lists/tables; clear chart titles/labels/axes and noncolor cues; tone, cultural/context fit and length; and easy-to-find what/who/when actions.

Review whether the theory-based findings test the actual ToC rather than recite a plan. Check the referenced pathway and documented revisions; per-link evidence sought/found and its meaning; original assumptions tested or unresolved; confirming/disconfirming evidence; plausible rivals such as other programs, background trends, selection or maturation; warranted ToC revisions; and causal language calibrated to support.

Review the contribution analysis as an evidence-based explanation of observed change, following a theory-based/contribution-analysis approach. Check whether program activity, context and co-factors form a coherent causal chain; rival explanations/background change and defensible counterfactual reasoning; bounded external co-factor contributions; calibrated strength/meaning of the contribution claim; and explicit source-to-inference steps a reader can retrace.

Review the supplied deliverable's ethical protections for its actual purpose, participant groups and stage. Request substantive text and relevant safe consent, data, risk, referral and oversight evidence; a plan or title is not implemented practice or approval.

Review GESI integration for the supplied deliverable's purpose, population and stage. Ask for actual text and safe supporting analyses, participation/disaggregation plans and safeguarding rules. Keep the deliverable, supporting records and evidence of practice distinct; attachments do not fill an undocumented section or prove implementation.

Review how the actual gender-relevant sections influence design and action beyond sex-disaggregated counts. Request relevant text, purpose/stage, wider gender analysis, methods and findings-to-action evidence where needed. Cite wider sources separately; they do not repair an undocumented section or prove safe implementation.

Review intersectional reasoning in the actual design or findings, using its questions, context/power analysis, intended comparisons, sampling/analysis and disclosure rules. Ask for missing substantive text; identity labels alone do not establish analysis. Keep identifiable participant or protected staff details out of AI input.

Review the documented relationship between positionality, reflexive practice and methodological decisions. Ask for actual relevant text, purpose/analytic approach, researcher-participant relationships and safe decision/reflection examples. A de-identified summary can suffice; do not demand private journals or sensitive identity disclosure.

Review whether the actual activity's protections address its participants and context. Request substantive protocols, participant risks, accessible consent/permission, staff responsibilities/competence, verified referral arrangements and oversight rules. Use safe summaries, not identifiable incident/participant records or protected staff details.

Review documented privacy controls for the actual data purpose, activity stage and applicable requirements. Request the relevant text, safe inventory/data-flow summaries, data categories and authorization, access/sharing/retention procedures, incident roles and policy/legal excerpts. Never request raw identifiable records, credentials, keys or exploitable configuration details.

Draft a context-specific do-no-harm protocol from the actual collection procedures/instruments, local risk assessment, participant needs, consent/data rules, staff competence/authority, schedule, safeguarding requirements and verified referral arrangements. Ask for missing essentials and give a provisional framework rather than a field-ready protocol. Keep identifiable incident and participant records out of the input.

Draft consent materials from the actual activity protocol and rules, participant needs/languages, recruitment, procedures/duration, risks/benefits/payments, data lifecycle/withdrawal limits, verified contact roles and review requirements. Ask for missing essentials; with unknown protocol or data-use facts give an explicitly incomplete outline and targeted questions, not participant-ready text. Input examples are not actual arrangements. Do not include participant identifiers.

Draft safeguarding procedures for the actual research/evaluation activity and population. Request substantive safeguarding/PSEA/child-protection policy sections, activity risks, consent/assent procedures, designated roles, verified reporting/referral directory and applicable duties. Policy or donor names alone do not establish rules. Unknown contacts, times and duties require safeguarding-lead resolution before use.

Review only survey-question wording, not response options, routing, structure or translation. Ask for actual item text/labels, shared instructions, intended construct/population, language and mode before item-specific findings; a survey name is insufficient.

Review the supplied answer choices in context of their questions, selection instructions, coding/routing notes, construct, population and mode. Request missing text; focus on options, not rewriting the whole instrument.

Review the actual translated instrument and, when available, its source side by side with item IDs, instructions, options and routing. Ask for intended language variety, population, construct definitions and review/pretest records. A language name or claimed process is not evidence. Without source text, limit conclusions to what the translation itself reveals and identify unavailable fidelity checks.

Review the actual FGD guide against its research topics, participant relationships, session time, consent and safety procedures. Ask for missing substantive text before guide-specific judgments; do not invent questions, safeguards or topic coverage.

Review the actual FGD probes with their parent questions, topics, participant context, timing and sensitive-topic instructions. Focus on probes rather than introduction, consent or role definitions. A guide name does not establish coverage.

Review the actual KII guide for the stated questions, informant roles, time, consent/recording and safety/data procedures. Request missing sections rather than crediting a title with their content.

Review the actual KII questions/probes for the stated research topics, informant roles and available interview time. Focus on the question set, leaving consent, recording and interviewer protocols to a separate review. Ask for missing text/context before item-specific judgments.

Review the actual FGD/KII protocol and guide together, including consent/recording scripts, purpose, participants, language needs and safeguarding/data policies. Request missing text; assess wraparound procedures as well as questions.

Review the supplied checklist and scoring instructions for its linked constructs, observation window, setting and paper/device mode. Ask for those details and available calibration/pilot records; a tool name cannot show its items.

Review the actual data-flow diagram or node-and-arrow narrative with linked schema/process descriptions, roles, access rules and staffing/connectivity context. Ask for missing evidence rather than inferring a working flow from system names.

Review the actual learning agenda against the relevant program theory, important uncertainties, intended decisions/users, available evidence and resource/timing limits. Request missing content before agenda-specific findings.

Review the actual assumptions alongside the ToC pathway, contextual evidence and proposed monitoring. Ask for these materials if absent; a ToC title cannot establish coverage or plausibility.

Review the actual ToC diagram and narrative against program scope and substantive contextual/causal evidence. Request missing content; a diagram title or reference list does not verify the pathway.

Review the actual ToC logic using its diagram/narrative, scope, assumptions and substantive evidence or accessible cited extracts. A ToC name or bibliography is not evidence that mechanisms work.

Review the actual roles matrix with relevant MEL workflows, organization/partner arrangements, delegated authority and available time/workload. Ask for these records before asserting capacity or accepted responsibilities from job titles.

Review how the actual plan, indicator sheet, findings brief or learning agenda links evidence to intended use. Ask for the deliverable, relevant decisions/users, definitions, reporting obligations and decision-rights context before document-specific judgments.

Review the supplied logframe assumptions with adjacent results and relevant ToC/risk context. Ask for actual text/item labels; a filename is insufficient. An isolated column supports wording review but not causal-link coverage.

Review the actual means-of-verification entries with indicator definitions, collection/reporting schedule, responsibilities and budget/system context. Ask for missing contents before document-specific judgments; an isolated source/platform name proves neither measurement fit nor feasibility.

Review the supplied evaluation questions against actual TOR purpose, users, scope, budget, team, timing, anticipated evidence and criteria. Ask for missing text; a title cannot establish answerability.

Review the actual evaluation budget/ceiling with TOR scope/methods, quantities/effort, schedule, currency/tax basis, dated cost evidence and procurement/funding rules. Request missing content before cost judgments; names are not rates or rules.

Review actual evaluation management text, mandate, decision rights and evaluation/procurement/COI rules. A committee name is not authority; request source sections before judging missing governance.

Review the actual inception engagement section, stakeholder map/safe summary, evaluation questions/timeline, access needs and disclosure/feedback arrangements. Request missing contents before judging participation; a title is not a plan.

Review the actual budget-versus-actuals narrative and linked figures for its intended decision. Ask for period, budget version, accounting/currency basis, explanations and variance/reallocation rules. Narrative alone supports wording/internal consistency, not verified reconciliation.

Review actual figures/images, labels/captions and source tables needed for value checks. Ask for contents, not report names. Descriptions support critique of described design/claims only; contrast, crowding, plotted values and actual scaling remain Not Assessable without readable figures.

Design transparent evaluator-selection criteria from actual TOR/questions/design needs, team model, budget, procurement/eligibility rules and diversity/access goals. Policy/funder names are not requirements. This creates criteria, not bidder rankings or an award.

Review actual inception workplan with TOR/deliverables, methods, team availability/effort, required reviews and known seasonal/access constraints. Ask for missing contents before deliverability judgments.

Reconcile actual conflicting monitoring records without altering raw evidence. Request safe competing extracts with stable labels/rows, definitions/inclusions, units/periods, coverage/versions and existing verification/decision rules. Names, assertions or unsigned proposals do not verify values. Use nonidentifying aggregates, not personal details or case histories; continue supported ledger entries.

Calculate reach only from comparable actual totals and overlap evidence. Request definition/eligibility, universe, window/cutoff, coverage, versions/replacements, units and repeats. Prefer permitted aggregate overlap; no direct identifiers. Minimal pseudonymous linkage results need confirmed permission here and remain sensitive; do not design linkage or promise weak-key deduplication.

Plan transition between actual dated indicator definitions without manufacturing continuity. Request old/new forms/codebooks, dates/reasons, coverage/history, baseline/target versions, permissible use and reporting/decision route. Use safe extracts, not personal records; absent essentials yield only an outline and requests.

Diagnose postcollection unit nonresponse and supported inference, not new sample sizing. Request purpose/estimand, population/design/selected counts, dispositions/eligibility/completion, safe frame/respondent summaries and documented weights. Use aggregates, not personal/contact records; missing essentials support conditional plans, not invented rates/bias.

Diagnose panel loss and comparability before paired change. Request target/paired quantity, cohort/eligibility, dates, validated linkage status/counts, dispositions, retained/lost baseline summaries, outcome changes, design/weights and replenishment. Use authorized safe aggregates/status summaries; no identifiers, contact records, keys or sensitive history. Unknown disclosure rules require review.

Document missing-data treatment for actual variables and the intended estimand/population, not a generic analysis workflow or claim of repaired data. Request codebook/skip/special-code rules, safe missingness counts/pattern denominators, unit/types, design/weights, models/indicator rules, timing/predictors and software/expertise constraints. Use authorized aggregates; no names, contacts, identifiers, keys or sensitive histories. Suppress unsafe groups or request disclosure review.

Assess missed preintervention measurement indicator by indicator, without inventing history. Request actual exposure/program timeline, constructs/populations/units/periods, intended comparison, dated candidate records/forms, coverage/permissions, resources and reporting rules. Missing essentials warrant only an options framework and focused requests, using safe extracts.

Assess secondary data for each actual new question/claim, separately from permission to use it. Request purpose/audience/estimand/population/unit/place/time, original collection/codebook, safe source content, versions/processing and use terms. No restricted personal records where metadata/aggregates suffice. Missing content supports a conditional plan, not findings or inferred permission.

Plan record linkage from safe schemas, not personal records or an executed join. Request purpose/output, actual permitted-use/access/sharing, definitions/units/versions/coverage, cardinality, aggregate key quality, authorized crosswalk description/validation and retention rules. Ask for no identifiers, key values/hashes, secrets/salts, contacts or sensitive histories. Unknown purpose/authorization blocks execution pending responsible review, while metadata planning may continue.

Draft an item-linked cognitive interview guide for actual survey text/IDs/instructions/options/routing, construct/use, population/language/mode/accessibility and session resources. Request missing safe text and consent/safeguarding/recording/data limits, not respondent answers/contacts. Without items/construct give only session structure, not invented questions or specific probes. This is response-process inquiry, not full pilot logistics or substantive program interviewing.

Plan transcript and translation quality checks before qualitative coding. Use the actual materials and permitted uses; do not conduct thematic analysis or invent participant findings.

Prepare an operational analysis handover from the actual artifacts so a successor can trace outputs and understand how to rerun the work.