harvestlisted

# Harvest Generate the three harvest outputs for a completed project and update Mnemosyne. **Law V:** Every completed project yields a blog post, a case study, and a testimonial request. These are not optional. ## Steps ### 1. Identify the project If invoked with a project name argument, use it. Otherwise read `{CONTENT_DIR}/portfolio/project-database.csv` and show the last 5 entries — ask which project to harvest. ### 2. Read the project record Read the row from `/media/data/dev/misc/upwork-proposals/context/portfolio/project-database.csv` for the named project. Extract all fields. Check current harvest status: - If all three harvest fields (Case Study, Blog Post, Testimonial Status) are already `published`, report harvest is complete and ask if the user wants to regenerate anything. - If some fields are done and some are not, note which are outstanding and continue with only those. ### 3. Gather context Read the following if they exist: - The project directory's `CLAUDE.md` — check CLAUDE.md's active projects table for the path - The project directory's `TODO.md` Do not invent details not present in the data or these files. ### 4. Generate the case study Create `{CONTENT_DIR}/portfolio/harvest/{slug}/case-study.md`: ``` # {Project Name} **Client:** {Client Name} · **Industry:** {Client Industry} · **Year:** {Year} **Stack:** {Tech Stack} ## The Brief {Project Description — expanded if context allows, otherwise verbatim} ## What We Delivered {Key Services ren