agentsclimarketplace
Sign in

Build case study

Skill Feedspace/Feedspace-Cookbook/plugins/feedspace-reviews/skills/build-case-study

Turn a customer testimonial into a structured case study draft - customer snapshot, challenge, solution, results, and a pull quote - grounded strictly in what the customer said, never inventing metrics or outcomes. Reviews come from a Feedspace account (via the Feedspace MCP), a CSV, or pasted text, so anyone can use it. Trigger when someone wants to build a case study, success story, or customer story from a review, testimonial, transcript, or feedback.From its SKILL.md

Install
npx -y skills add Feedspace/Feedspace-Cookbook --skill build-case-study

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.9 KB, 800 tokens by cl100k_base, as published. Nobody here has run it

Build case study

Turn one strong testimonial (or a few reviews from the same customer) into a clean, structured case study draft: a customer snapshot, the challenge, the solution, the results, and a pull quote. Every line is grounded in what the customer actually said - no invented metrics, timelines, or outcomes. The source can be a Feedspace account through the Feedspace MCP, a CSV, or pasted text.

The full method lives in references/case-study-method.md. Follow it exactly. It runs entirely in this session and needs no API key; pulling reviews from the Feedspace MCP is a separate, read-only connection.

Steps

  1. Get the source testimonial.

    • Feedspace MCP: first check whether it is already connected - are Feedspace tools (for example list_workspaces) available? If yes, just use them; do not ask the user to reconnect. Find a rich candidate (a detailed video or audio testimonial, or a long text review - filter_reviews by type and 4-5 stars, or search by the customer's name), then call get_review for the full detail, including the transcript and the reviewer's name, role, and company. Only if the tools are not available, point the user to setup in references/mcp-source.md, which also covers how to find candidates and where each field lives.
    • Otherwise read a CSV (see references/csv-format.md) or ask the user to paste the testimonial plus any known customer details (name, role, company).

  2. Run a completeness check, then confirm the subject. A case study rests on four pillars: who the customer is, the challenge they faced, the solution, and the result. Tell the user which pillars the source actually supports and which are missing, so nothing gets invented to fill a gap. If the source has several different customers, build one at a time - ask which, or offer the richest.

  3. Build the case study by following references/case-study-method.md: headline options, customer snapshot, challenge, solution, results, and one verbatim pull quote - each traceable to the source, in the customer's own language.

  4. Offer one sensible next step, then stop. Options: draft a short outreach message asking the customer to fill a missing pillar (for example the before-state or a metric), produce a shorter social-proof pull-quote version, or build a case study from a different testimonial. If the review came from the Feedspace MCP, you can also pull another candidate over the live connection.

Guardrails

  • Ground everything in the source. Never invent a challenge, a number, a timeline, a company detail, or an outcome the customer did not state. If a pillar is missing, say so - do not manufacture it.
  • Quote the pull quote verbatim and attribute it. Do not turn a compliment into a metric or a feature into a result.
  • If the source is too thin for a real case study (a one-liner like "Great product, love it"), say so and offer a short testimonial highlight instead, plus what to ask the customer for.
  • Detect the language of the testimonial and write the case study in it.
  • This skill only reads and drafts. The output is a draft for the user to review, edit, and publish themselves. The Feedspace MCP path is read-only - never modify, favourite, feature, or delete anything.

What ships with it: 5 files

20.8 KB alongside SKILL.md

references/

Gives 0 of the 12 instructions most learn study skills give in 800 tokens

Counted across 546 of the 573 authors here whose files we hold, read 2026-08-07

  • Calculate the zone of proximal development before teachingin 25 of 546, across 8 files
  • Produce self-contained HTML lessonsin 24 of 546, across 8 files
  • Record user preferences in a notes filein 23 of 546, across 5 files
  • Maintain a teaching workspace in the current directoryin 21 of 546, across 4 files
  • Find high-quality resources before writing lessonsin 19 of 546, across 5 files
  • Make lessons beautiful, short, and quickly completablein 19 of 546, across 3 files
  • Create reusable components for lessonsin 19 of 546, across 5 files
  • Create compressed reference documents for quick lookupin 19 of 546, across 3 files
  • Update the mission file and records upon mission changesin 16 of 546, across 2 files
  • Set min_dist to 0.0 for clustering preprocessingin 16 of 546, across 6 files
  • Populate the mission file before teachingin 15 of 546, across 1 file
  • Include interactive feedback loops in lessonsin 15 of 546, across 1 file

Said here and by no other author read

  • turn one testimonial into a structured case study
  • follow the method file exactly
  • check existing MCP connection before asking user
  • filter reviews to find a rich candidate
  • tell the user which pillars are missing
  • build one case study at a time

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,645. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.