Price Check: Market Research for Purchasing Intake

PREREQUISITE: Load the following skills to execute this recipe: reverb-listings

Research active listings and recent sold prices on Reverb.com to establish a used-market price range for gear submitted for possible purchase.

Required Output Format

You MUST produce two markdown tables — one for sold listings, one for active listings. Do not produce prose summaries instead of tables.

Sold Listings

Title	Condition	Caveats	Asking	Sold	Listed	Source
[title]	[condition]	[repairs/mods or —]	[original_price.display]	[price.display]	[created_at date]	link

Active Listings

Title	Condition	Caveats	Asking	Listed	Source
[title]	[condition]	[repairs/mods or —]	[price.display]	[created_at date]	link

Source column rules:

• Every row MUST have a hyperlinked source using _links.web.href
• Sold listing URLs: append ?show_sold=true
• Format: [View](https://reverb.com/...)

After the tables, add a Price Range Summary (2–4 sentences max): sold baseline range by condition, active ask range, notable discount depth if present.

---

Steps

1. Fetch active and sold listings in parallel (write to files to avoid output truncation):

   revcli listings list --query "MAKE MODEL" --page-all --per-page 50 --page-limit 4 --format json --output /tmp/active.json
   revcli listings list --query "MAKE MODEL" --params '{"show_only_sold":true}' --page-all --per-page 50 --page-limit 4 --format json --output /tmp/sold.json
   

2. Parse both files with Python. --page-all emits one JSON object per page, not a single array — use raw_decode:

   import json
   from json import JSONDecoder
   
   def load_pages(path):
       with open(path) as f:
           text = f.read().strip()
       decoder, pos, listings = JSONDecoder(), 0, []
       while pos < len(text):
           while pos < len(text) and text[pos] in ' \n\r\t':
               pos += 1
           if pos >= len(text):
               break
           obj, pos = decoder.raw_decode(text, pos)
           listings.extend(obj.get('listings', []))
       return listings
   
   active = load_pages('/tmp/active.json')
   sold   = load_pages('/tmp/sold.json')
   

3. Extract per listing, using python:

   • title
   • condition.display_name — e.g. "Excellent", "Brand New"
   • price.display — asking or final sold price
   • original_price.display — pre-discount price (sold listings only; gap = negotiability signal)
   • created_at — date listed (sort by this, descending)
   • description — HTML; strip tags; scan for caveats (replaced parts, repairs, mods, damage); do NOT flag negated phrases ("no cracks", "not modified")
   • _links.web.href — listing URL (required for Source column)

4. Apply filters before building tables:

   • Exclude condition.display_name == "Brand New" from used-market baseline
   • Flag but include listings with repairs/mods/replaced parts — put the caveat in the Caveats column
   • Flag but include "unplayed"/"new-old-stock" listings — treat as Excellent condition, note in Caveats

5. Populate the tables defined in Required Output Format above, then write the summary.