Amazon-domain general analysis and multi-endpoint research

ZooData — Amazon Seller Data Analysis

by SerendipityOneInc · MIT license · ★ 74 on GitHub · updated Sep 22

Amazon-domain general analysis and multi-endpoint research engine. Handles broad or composite Amazon research requests that span multiple data dimensions or have no single specialized angle.

---
name: amazon-analysis
description: >
  Amazon-domain general analysis and multi-endpoint research engine.
  Handles broad or composite Amazon research requests that span multiple data
  dimensions or have no single specialized angle. For focused category-market
  discovery, market-entry evaluation, or category trends, use
  amazon-market-analysis.
  Use when:
  - user asks for multi-endpoint Amazon research or composite reports
  - user asks "what kind of Amazon analysis can I run" or wants an overview
    of available Amazon insights
  - user wants broad Amazon data exploration with no single specific
    deliverable in mind
  Uses {skill_base_dir}/scripts/zoodata.py. Requires ZOODATA_API_KEY.
metadata:
  version: "1.1.13"
  author: SerendipityOneInc

Overview

Respond in the user's language. Use this skill for broad or composite Amazon research without a single specialized deliverable. Route a focused market-entry, market-discovery, or category-trend request to amazon-market-analysis; route a focused keyword, listing, pricing, review, or competitor request to its dedicated skill.

What you get

Networkthe bundled CLI sends authenticated requests to https://api.zoodata.ai and refuses an untrusted ZOODATA_BASE_URL.
ExecutionPython 3 standard-library shared CLI at {skill_base_dir}/scripts/zoodata.py.
Local filesreads optional ~/.zoodata/config.json; review fallback may use a private temporary directory.
Sent to the APIkeywords, category paths/IDs, ASINs, marketplace, dates, and numeric filters. Seller profile text remains local.
Creditsevidence calls consume account credits; the credential-only check path does not.

Manual install (npx, ZIP)

Claude Codenpx skills add SerendipityOneInc/ZooData-Skills --skill amazon-analysis -g -a claude-code

Claude Code (manual)

mkdir -p ~/.claude/skills && curl -sL https://codeload.github.com/SerendipityOneInc/ZooData-Skills/tar.gz/7efe44572d17 \
  | tar -xz -C ~/.claude/skills --strip-components=1 ZooData-Skills-7efe44572d17/amazon-analysis

Windows (PowerShell)

New-Item -ItemType Directory -Force "$HOME\.claude\skills" | Out-Null; Invoke-WebRequest "https://m.agentalley.io/zip/827532b7124295277bb7/amazon-analysis.zip" -OutFile "$env:TEMP\amazon-analysis.zip"; Expand-Archive "$env:TEMP\amazon-analysis.zip" -DestinationPath "$HOME\.claude\skills" -Force

Claude.ai / desktopDownload the ZIP and upload it under Settings → Capabilities → Skills.

Codex, Cursor, other agentsPoint them at SKILL.md. The other files are plain text.

Connects with Agent Alley MCP · agentalley.io/mcp

Files included

SKILL.md5.7 KB
.clawhubignore12 B
README.md3.0 KB
references/9 files · 77.7 KB
scripts/2 files · 187.3 KB

Read the files here

SerendipityOneInc/7efe44514 files

SKILL.md

Show the full text73 lines

amazon-analysis/SKILL.md73 lines · 5.7 KB

PreviewCode

Outline

RawView on GitHub

ZooData — Amazon Seller Data Analysis

Respond in the user's language. Use this skill for broad or composite Amazon research without a single specialized deliverable. Route a focused market-entry, market-discovery, or category-trend request to amazon-market-analysis; route a focused keyword, listing, pricing, review, or competitor request to its dedicated skill.

Start here

Classify the user's question. For a single data point, use the quick path in references/execution-guide.md; for multi-endpoint research, load its full workflow.
Read references/cli-contract.md before selecting or invoking a command. Inspect the bundled CLI's top-level and selected subcommand help.
Read references/reference.md for endpoint fields and local CLI options. Load the applicable scenario module: scenarios-composite.md for broad cross-validation, scenarios-eval.md for product evaluation, scenarios-expand.md for expansion, scenarios-listing.md for listing context, scenarios-ops.md for operations, or scenarios-pricing.md for pricing context.
Apply the shared interpretation and reporting rules in references/execution-guide.md; use the selected scenario for its specific evidence and conclusion scope.

Source-of-truth boundaries

This SKILL.md owns trigger routing, module loading, the responsibility map, and non-negotiable runtime boundaries. It does not define product presets, thresholds, endpoint fields, procedures, or report templates.
references/cli-contract.md owns shared CLI invocation, command identity, result classification, retries, and terminal or partial-result handling.
references/reference.md owns endpoint availability, request and response fields, CLI parameter mapping, data shapes, and field identity. It does not set analytical thresholds.
references/execution-guide.md owns shared workflow steps, product-preset selection, market-health interpretation, cross-endpoint analysis, and user-facing output rules. It does not redefine endpoint fields.
Scenario modules own the evidence selection, conclusions, and scenario-specific report sections for their named task. They may narrow shared workflow rules but cannot redefine endpoint contracts or shared output rules.
{skill_base_dir}/scripts/zoodata.py owns local preset expansion, request construction, transport metadata, and credential handling; it does not issue business conclusions.

Shared CLI Contract

Before selecting or invoking the first command, read and apply the local references/cli-contract.md. Reapply it after every granular or composite result and before any fallback, additional call, state write, interpretation, or report. The local references/execution-guide.md owns user-facing failure rendering.

Local Interface Failure Output

Use the failure rendering in references/execution-guide.md; direct a user with exhausted credits to https://zoodata.ai/en/pricing.

Non-negotiable boundaries

Require ZOODATA_API_KEY; use the bundled CLI and the local command manifest. The credential-only check path precedes evidence calls. Stop and follow the shared CLI contract on credential, credit, validation, or terminal interface failures; do not invent missing evidence.
Resolve and preserve a category path or ID before category-scoped comparisons. Distinguish inferred category paths from direct matches. Preserve marketplace, date, category scope, sample type, and the difference between full-category marketTotal and selected Top 100 marketSample metrics.
Treat seller budget, experience, risk tolerance, and other profile text as local interpretation inputs. Send only the documented category/product identifiers and numeric filters to the API.
State estimated credit cost before a broad multi-call scan. Require the user's explicit monitoring request before creating recurring work or saved baselines.

Capabilities and data flow

Network: the bundled CLI sends authenticated requests to https://api.zoodata.ai and refuses an untrusted ZOODATA_BASE_URL.
Execution: Python 3 standard-library shared CLI at {skill_base_dir}/scripts/zoodata.py.
Local files: reads optional ~/.zoodata/config.json; review fallback may use a private temporary directory.
Sent to the API: keywords, category paths/IDs, ASINs, marketplace, dates, and numeric filters. Seller profile text remains local.
Credits: evidence calls consume account credits; the credential-only check path does not.

Execution entry

python {skill_base_dir}/scripts/zoodata.py <documented-subcommand> ...

This skill allows categories, market, market-structure-profile, market-history, products, competitors, product, analyze, report, opportunity, history, brand-overview, brand-detail, price-band-overview, price-band-detail, check, reviews-raw, review-tag-prompt, review-reduce-prompt, and review-aggregate.

1	---
2	name: amazon-analysis
3	description: >
4	Amazon-domain general analysis and multi-endpoint research engine.
5	Handles broad or composite Amazon research requests that span multiple data
6	dimensions or have no single specialized angle. For focused category-market
7	discovery, market-entry evaluation, or category trends, use
8	amazon-market-analysis.
9	Use when:
10	- user asks for multi-endpoint Amazon research or composite reports
11	- user asks "what kind of Amazon analysis can I run" or wants an overview
12	of available Amazon insights
13	- user wants broad Amazon data exploration with no single specific
14	deliverable in mind
15	Uses {skill_base_dir}/scripts/zoodata.py. Requires ZOODATA_API_KEY.
16	metadata:
17	version: "1.1.13"
18	author: SerendipityOneInc
19	homepage: https://github.com/SerendipityOneInc/ZooData-Skills
20	openclaw: {"requires": {"env": ["ZOODATA_API_KEY"]}, "primaryEnv": "ZOODATA_API_KEY"}
21	---
22
23	# ZooData — Amazon Seller Data Analysis
24
25	Respond in the user's language. Use this skill for broad or composite Amazon research without a single specialized deliverable. Route a focused market-entry, market-discovery, or category-trend request to `amazon-market-analysis`; route a focused keyword, listing, pricing, review, or competitor request to its dedicated skill.
26
27	## Start here
28
29	1. Classify the user's question. For a single data point, use the quick path in `references/execution-guide.md`; for multi-endpoint research, load its full workflow.
30	2. Read `references/cli-contract.md` before selecting or invoking a command. Inspect the bundled CLI's top-level and selected subcommand help.
31	3. Read `references/reference.md` for endpoint fields and local CLI options. Load the applicable scenario module: `scenarios-composite.md` for broad cross-validation, `scenarios-eval.md` for product evaluation, `scenarios-expand.md` for expansion, `scenarios-listing.md` for listing context, `scenarios-ops.md` for operations, or `scenarios-pricing.md` for pricing context.
32	4. Apply the shared interpretation and reporting rules in `references/execution-guide.md`; use the selected scenario for its specific evidence and conclusion scope.
33
34	## Source-of-truth boundaries
35
36	- This `SKILL.md` owns trigger routing, module loading, the responsibility map, and non-negotiable runtime boundaries. It does not define product presets, thresholds, endpoint fields, procedures, or report templates.
37	- `references/cli-contract.md` owns shared CLI invocation, command identity, result classification, retries, and terminal or partial-result handling.
38	- `references/reference.md` owns endpoint availability, request and response fields, CLI parameter mapping, data shapes, and field identity. It does not set analytical thresholds.
39	- `references/execution-guide.md` owns shared workflow steps, product-preset selection, market-health interpretation, cross-endpoint analysis, and user-facing output rules. It does not redefine endpoint fields.
40	- Scenario modules own the evidence selection, conclusions, and scenario-specific report sections for their named task. They may narrow shared workflow rules but cannot redefine endpoint contracts or shared output rules.
41	- `{skill_base_dir}/scripts/zoodata.py` owns local preset expansion, request construction, transport metadata, and credential handling; it does not issue business conclusions.
42
43	## Shared CLI Contract
44
45	Before selecting or invoking the first command, read and apply the local `references/cli-contract.md`. Reapply it after every granular or composite result and before any fallback, additional call, state write, interpretation, or report. The local `references/execution-guide.md` owns user-facing failure rendering.
46
47	### Local Interface Failure Output
48
49	Use the failure rendering in `references/execution-guide.md`; direct a user with exhausted credits to https://zoodata.ai/en/pricing.
50
51	## Non-negotiable boundaries
52
53	- Require `ZOODATA_API_KEY`; use the bundled CLI and the local command manifest. The credential-only `check` path precedes evidence calls. Stop and follow the shared CLI contract on credential, credit, validation, or terminal interface failures; do not invent missing evidence.
54	- Resolve and preserve a category path or ID before category-scoped comparisons. Distinguish inferred category paths from direct matches. Preserve marketplace, date, category scope, sample type, and the difference between full-category `marketTotal` and selected Top 100 `marketSample` metrics.
55	- Treat seller budget, experience, risk tolerance, and other profile text as local interpretation inputs. Send only the documented category/product identifiers and numeric filters to the API.
56	- State estimated credit cost before a broad multi-call scan. Require the user's explicit monitoring request before creating recurring work or saved baselines.
57
58	## Capabilities and data flow
59
60	- Network: the bundled CLI sends authenticated requests to `https://api.zoodata.ai` and refuses an untrusted `ZOODATA_BASE_URL`.
61	- Execution: Python 3 standard-library shared CLI at `{skill_base_dir}/scripts/zoodata.py`.
62	- Local files: reads optional `~/.zoodata/config.json`; review fallback may use a private temporary directory.
63	- Sent to the API: keywords, category paths/IDs, ASINs, marketplace, dates, and numeric filters. Seller profile text remains local.
64	- Credits: evidence calls consume account credits; the credential-only `check` path does not.
65
66	## Execution entry
67
68	```bash
69	python {skill_base_dir}/scripts/zoodata.py <documented-subcommand> ...
70	```
71
72	This skill allows `categories`, `market`, `market-structure-profile`, `market-history`, `products`, `competitors`, `product`, `analyze`, `report`, `opportunity`, `history`, `brand-overview`, `brand-detail`, `price-band-overview`, `price-band-detail`, `check`, `reviews-raw`, `review-tag-prompt`, `review-reduce-prompt`, and `review-aggregate`.
73