# Evaluation planning starter pack

This pack helps you complete one M&E job with free instructions and fictional attachments. Your Claude account has its own access and usage limits. Checked in Claude Code. Anthropic's help page says Free and paid plans can use uploaded skills once code execution is turned on; we checked browser upload on a paid plan only. No API key or site signup is needed.

## Inputs and start

Supply `evaluation-planning.md` or your equivalent readable files. examples/normal and examples/difficult are separate fictional cases; never combine them. Browser users upload files; Code users give explicit file paths or a scoped folder. If extraction fails, provide a readable text export. This pack promises Markdown output, not Word, PDF or spreadsheet rendering.

Enable these workspace skills: evaluation-questions, question-method-alignment, evaluation-tor-draft. Also enable the review skill(s) tor-review from me-review. In Code load both plugins and use `/me-workspace:<skill>` and `/me-review:<review>`. In a browser upload individual skill packages and name the skill. Copy/paste fallback: paste each relevant SKILL.md in sequence with the files or text it needs; browser behavior still needs separate verification.

Starter: Complete evaluation planning with my supplied documents. Use evaluation-questions to write questions.md, question-method-alignment to write matrix.md, and evaluation-tor-draft to write tor-draft.md. Use me-review:tor-review to write a prioritized review.md of that draft against the supplied constraints. Revise into tor-revised.md and record each material finding, change or unresolved decision in change-record.md. Then synchronize questions.md and matrix.md to the final TOR question IDs, wording, core/optional status, priorities, day allocations and schedule. Do not leave conflicting versions for the user to reconcile. Preserve the original TOR draft separately for comparison. Check every final question has an evidence/method row and all interview and fieldwork dates fall within the supplied fieldwork window, if one is given; distinguish later desk analysis from fieldwork. If a day cap is given, check task days add to it. Keep unresolved choices explicit. No conditional source may become confirmed access. Return the final artifacts in editable Markdown. Separate facts, assumptions and proposed choices. Make unresolved decisions explicit. Treat core/optional status, day allocations and intermediate dates as proposals with a stated basis. If no defensible estimate is possible, leave it unresolved rather than inventing effort or commitments.

## Save and resume

Number decisions with the stage prefix (EVAL-D1, EVAL-D2) so each ID means one decision across all files. Save intermediate outputs as you go. If interrupted, use handover-template.md to state the input files/source IDs, completed files, decisions/assumptions, remaining steps and exact restart instruction. In a fresh session provide the original input, completed artifacts and handover. Re-validate inherited facts and status claims against the original sources before copying them forward. A handover is a navigation aid, not authority for approval, consent or measurement definitions. Use three evidence states: explicitly confirmed, explicitly contradicted, or unconfirmed. No supporting record means unconfirmed; it cannot justify keeping a negative status claim. If a resumed claim has no supporting record, mark it unknown rather than keeping it. Distinguish approval status of pre-existing program items from recommendations created in this draft (proposed). Do not rely on earlier chat memory or undocumented author corrections.

Expected files: questions.md, matrix.md, tor-draft.md, review.md, tor-revised.md, change-record.md. Check that you can open each file, follow source IDs back to inputs, distinguish proposals from commitments, and see the next decision. Human usability validation remains pending.