Evaluation Questions
Prioritize answerable questions tied to the decisions an evaluation must support.
Task Skill
Checked in Claude Code
Inputs, Examples and Files for Evaluation QuestionsAI Skills
Practical M&E skills, document reviews and workflows to use in Claude.
Try evaluation questions, indicators or logframe review.
Find an M&E Task
Choose a task to see what to provide, what you can expect and how to use it. 12 task skills, 12 document reviews and 3 workflow packs.
Use these resources in your own Claude account. Check drafts and reviews against your source documents before sharing them. Checking Evidence and Limits
6 matching entries
Question and Method Alignment
Match evaluation questions to feasible evidence, methods and limitations.
Task Skill
Checked in Claude Code
Inputs, Examples and Files for Question and Method AlignmentEvaluation TOR Draft
Draft a source-linked TOR with feasible scope and unresolved decisions visible.
Task Skill
Checked in Claude Code
Inputs, Examples and Files for Evaluation TOR DraftSampling Plan Draft
Draft a sampling plan with visible calculations, or check the sizing and selection of an existing one.
Task Skill
Checked in Claude Code
Inputs, Examples and Files for Sampling Plan DraftBaseline Planning
Plan baseline measurement for every indicator, or check a baseline against the framework, without inventing starting values.
Task Skill
Checked in Claude Code
Inputs, Examples and Files for Baseline PlanningEvaluation Planning
Move from evaluation decisions to questions, methods and a revised TOR.
Workflow Pack
Checked in Claude Code
Inputs, Examples and Files for Evaluation PlanningSee an Example Review
A hand-written fictional illustration of Logframe Review. This is not an observed test result or a complete review.
Fictional Review Excerpt
Fictional excerpt for explanation only. This is not an observed test result or a complete review.
A fictional framework lists training sessions as its outcome, with attendance as the only indicator. No adoption measure is supplied.
Illustrative Review Note
Fictional excerpt for explanation only. This is not an observed test result or a complete review.
The supplied outcome and indicator describe delivery, not a demonstrated change in practice. Ask the program team to clarify the intended outcome and its evidence; do not invent an adoption target.
Choose How to Use AI Skills
Claude Code is the route checked across the recorded task cases. Browser upload was checked for one skill on a paid plan. Check Claude’s documentation for current account requirements and setup steps.
Claude Code
Add the marketplace, then install the review and workspace plugins using the recorded setup commands:
/plugin marketplace add mestudiohq/me-review
/plugin install me-review@me-review
/plugin install me-workspace@me-reviewIf the install summary asks you to reload plugins, follow that instruction. Then type /me-workspace: or /me-review: to see the skills. Each task page gives its command and required inputs.
Claude in Your Browser
Open a task page and check its individual skill files. When a verified skill ZIP is available, that page provides the download.
The recorded browser check covered Evaluation Questions on a paid plan. It does not establish that every skill or account type behaves the same way.
Check the provider’s current instructions before choosing an account or uploading files. Follow your organization’s data rules.
Markdown is the checked output format. Word, PDF and spreadsheet output are not promised. Each method has its own review rules; an AI review is not approval.
Guides
Getting Started
Install the plugins, choose a job, start it, check the Markdown file you get back, and pause or resume with a handover.
A Worked Project
One fictional program taken through evaluation questions, a feasible measurement plan and a report with proposed actions.
Usage and Credit
How to credit M&E Studio when you share or adapt these guides, and how the plugin license differs.
Checking Evidence and Limits
Checks ran in Claude Code on fictional cases, including incomplete and unsuitable inputs, the three workflows, and pausing and resuming. A second model checked the outputs against expected findings. Earlier failures were fixed and the affected cases repeated.
Browser upload passed for Evaluation Questions on a paid plan. Free accounts, copy and paste, and practitioner feedback were not checked. Of the 12 document reviews, 4 were included in the latest recorded checks; the others come from earlier releases.
The examples are fictional and written by hand. They are not real client work or observed test output. Review every output against your supplied evidence before use.
Source revision 18dd509 · Checked 2026-10-26