v1.2.2 — Realtime Reviews Fallback + SKILL.md Spec Compliance

This release rolls up 11 PRs since v1.2.0. Two major themes: a self-contained review fallback toolkit when /reviews/analysis lacks aggregation, and SKILL.md name: field compliance for strict skill loaders (Codex).

Highlights

Realtime Reviews Fallback Toolkit (#57)

• New endpoint integration: /openapi/v2/realtime/reviews — cursor-paginated raw review fetch (10/page, max 100, 1 credit/page, US+UK)
• New CLI commands in apiclaw.py: reviews-raw, review-tag-prompt, review-reduce-prompt, review-aggregate — prompt-as-data Map/Reduce toolkit that uses the caller's own LLM (no external dependency)
• All 8 review-using SKILL.md files document the fallback chain self-contained
• Reference docs updated with realtime/reviews and reviews/search schemas

SKILL.md name Field Spec Compliance (#65)

All 10 SKILL.md name: fields rewritten to kebab-case matching the parent directory name, per the Agent Skills open standard. Codex strict loader was rejecting skills because of this — now they load.

• ClawHub: URL, slug, install directory, page display all unchanged. openclaw skills update will refuse on fingerprint mismatch — pass --force.
• Claude Code: no user-visible change.
• Codex: skills now load. npx skills add users with old installs will have orphaned directories — cleanup command below.

Other (#51–#56, #62, #63)

• Sync infrastructure: scripts/sync-scripts.sh repointed at canonical-source banner (#63); SKILL.md PR CI checks for path prefix (#53)
• Bug fixes: market-entry keyword-only fallback when categoryPath empty (#62); API v2 field renames + rate limiting + category auto-detection (#52)
• ClawHub: security scan warnings resolved (#55); patch version bumps for republish (#56)
• Docs: README skill tables updated (#51, #54)

CI Removed

check-skill-name-unchanged job removed. Its premise (that name: determines installed directory) was incorrect — ClawHub uses its own slug fixed at publish time, and other install paths are independent.

Migration

If you installed via npx skills add before v1.2.2 (and saw the long human-readable name as your slash command), you'll have orphaned skill directories. Clean them up:

npx skills list
npx skills remove "amazon-daily-market-radar-automated-monitoring-alerts"
npx skills remove "amazon-review-intelligence-extractor-consumer-insights-from-1b-reviews"
# (and any other long-name orphans shown by `list`)
npx skills add SerendipityOneInc/APIClaw-Skills

ClawHub and git clone users: no action needed.

Per-skill versions in this release

skill	version
apiclaw	1.1.3
amazon-analysis	1.1.7
amazon-competitor-intelligence-monitor	1.1.3
amazon-daily-market-radar	1.0.3
amazon-listing-audit-pro	1.0.3
amazon-market-entry-analyzer	1.0.3
amazon-market-trend-scanner	1.0.2
amazon-opportunity-discoverer	1.0.3
amazon-pricing-command-center	1.1.2
amazon-review-intelligence-extractor	1.0.3

See CHANGELOG.md for full details.

v1.2.0 — API V2 Migration, 10 Skills, New Skill Format

What's New

📦 API V2 Migration

All 10 skills updated to APIClaw API V2 — new field naming conventions, 3 endpoint path changes, and pageSize limit increased to 100. Added live OpenAPI spec reference to all skills for self-service field verification.

🏗️ 10 Skills — Unified & Streamlined

Foundation (2):

Skill	What It Does	Input	Output
APIClaw — Amazon Commerce Data, 11 Endpoints	Direct access to all 11 API endpoints	Keyword, category, ASIN, or brand	Raw API data with field mapping
Amazon Analysis — Full-Spectrum Research & Seller Intelligence	13 selection modes + unified research workflows	Keyword/category/ASIN + intent	Analysis findings, top products, confidence-tagged insights

Specialized (8):

Skill	What It Does	Input	Output
Amazon Competitor Intelligence Monitor	Dual-mode competitive intelligence with tiered alerts	Keyword or ASIN(s)	Competitor matrix, scores (1-100), three-tier alerts
Amazon Daily Market Radar — Automated Monitoring & Alerts	Daily monitoring — price, BSR, competitors, reviews	Your ASINs (1-10) + keyword	RED/YELLOW/GREEN alerts, KPI dashboard, action items
Amazon Listing Audit Pro — 8-Dimension Health Check	Listing quality audit with optimization recommendations	Your ASIN + keyword	Score (X/100, A-F), 8-dimension scorecard, priority fixes
Amazon Market Entry Analyzer — GO/CAUTION/AVOID Verdicts	One-click market viability assessment	Keyword or category path	Verdict, sub-market landscape, entry strategy
Amazon Market Trend Scanner — Daily Category Radar ✨	Category landscape scanning and trend discovery	1+ category paths or keywords	Trend dashboard, Hot Categories TOP 5, risk alerts
Amazon Opportunity Discoverer — Niche Scanner & Scoring	Profile-driven opportunity scanning with 7-dimension scoring	Budget + experience + keyword	Top 10 opportunities (S/A/B/C), risk alerts
Amazon Pricing Command Center — RAISE/HOLD/LOWER Signals ✨	Data-driven pricing signals, auto-detects leaf category	One or more ASINs	Price signal, price band heatmap, BuyBox analysis
Amazon Review Intelligence Extractor — Consumer Insights from 1B+ Reviews	Deep consumer insights across 11 analysis dimensions	ASIN(s) or category keyword	Pain points, buying factors, differentiation roadmap

✏️ Skill Changes

• Renamed: Dynamic Pricing Intelligence Agent → Amazon Pricing Command Center — RAISE/HOLD/LOWER Signals
• Renamed: Market Radar — Daily Trend Scanner → Amazon Market Trend Scanner — Daily Category Radar
• Removed: Amazon Blue Ocean Finder (merged into Opportunity Discoverer)
• Disabled: beginner selection mode — now 13 modes (was 14)
• Streamlined: Pricing skill — removed cost/COGS input, focused on market-data-driven signals

🐛 Bug Fix

• Fixed argparse prefix matching bug where --page 1 silently overwrote --page-size value

📋 Repo Improvements

• Unified all reference.md to complete 11-endpoint version (single source of truth)
• Added live OpenAPI spec reference to all SKILL.md
• README (EN + CN) updated with new input/output skill format
• 41 unit tests passing

Full Changelog: v1.1.4...v1.2.0

v1.1.4 — 7 Hero Skills, 11 Endpoints

What's New

🚀 7 New Hero Skills (v1.0.0)

Purpose-built skills for specific Amazon seller workflows:

• Amazon Competitor War Room — Competitive intel & battle strategy
• Amazon Daily Market Radar — Automated monitoring & alerts
• Amazon Listing Audit Pro — 8-dimension health check
• Amazon Market Entry Analyzer — GO/CAUTION/AVOID verdicts
• Amazon Opportunity Discoverer — Niche scanner & scoring
• Amazon Pricing Command Center — Optimal price positioning
• Amazon Review Intelligence — Consumer insights from 1B+ reviews

📦 APIClaw v1.1.0

• Expanded from 6 to 11 API endpoints: added price-band overview/detail, brand overview/detail, product history
• Complete endpoint documentation rewrite

🎯 Amazon Analysis v1.1.4

• Major SKILL.md rewrite: improved intent routing, workflow structure, agent instructions
• New references/execution-guide.md — step-by-step execution playbook
• Updated API reference for all 11 endpoints

📋 Repo

• Added scoring-methodology.md — unified quality scoring framework
• README restructured: 9 skills with tiered layout, complete project structure
• Unified skill names with descriptive subtitles
• CHANGELOG updated

Full Changelog: v1.1.3...v1.1.4

v1.1.3 — Security Hardening + Output Standards

What's Changed

🔒 Security

• Fixed potential API key exposure risk in example configurations
• Removed hardcoded endpoint URLs from skill documentation

📝 Quality

• Rewrote all reference files with improved field descriptions and usage examples
• Improved SKILL.md descriptions for better agent intent matching
• Enforced mandatory API usage tracking in output
• Enforced mandatory Data Source block in Full Mode output
• Added mandatory pre-execution checklist for Full Mode
• Optimized name and description for search discoverability

Full Changelog: v1.1.2...v1.1.3

v1.1.2 — Credits Tracking + Review Analysis + API Field Updates

What's New

🆕 New Features

• Credits Tracking: Every API response now includes creditsConsumed and creditsRemaining
• Realtime Data Supplementation: In Full-mode analysis, automatically calls realtime/product for top 3-5 ASINs
• Reviews/Analyze Endpoint: New analyze command for AI-powered review analysis (11 dimensions)
• New Scenario 4.6: Category Consumer Insights

🔄 API Field Updates (Breaking)

• review → rating rename across fields, filters, and CLI args
• topReviews removed from realtime/product (use reviews/analyze instead)
• sortBy updated: reviewCount → ratingCount

📊 Market Interface — 6 New Product Fields

• sampleNewSkuAvgPrice, sampleNewSkuAvgMonthlySaleCnt, sampleNewSkuAvgMonthlySaleAmt, sampleNewSkuAvgRatingAmt, sampleNewSkuAvgRatingCnt, sampleNewSkuCount

📦 Housekeeping

• Slimmed SKILL.md from 448 → 417 lines
• Initial release of general apiclaw/ skill (platform overview, 6 endpoints)

Full Changelog: v1.1.1...v1.1.2

v1.1.1 — Credential Security Enhancement

What's Changed

🔒 Security

• Improved credential handling: require user confirmation before writing to config.json
• Explicitly warn users that writing secrets to disk is not recommended
• Emphasize environment variable as the preferred method

📦 Metadata

• Use ClawHub standard metadata format for credential declaration
• Replace custom 'credentials' field with metadata.openclaw.requires.env + primaryEnv

Full Changelog: v1.1.0...v1.1.1

v1.1.0 — Security Fixes + Listing Module + Mode Corrections

What's New

🔒 Security

• Removed API key logging from cmd_check() (resolves CodeQL CWE-312/359/532)
• Added author, homepage, credentials metadata to SKILL.md frontmatter
• Updated credential docs: environment variable as preferred method
• Added SECURITY.md

🐛 8 Documentation Fixes

• Added interface data differences table (realtime vs offline field mapping)
• Fixed multi-product comparison: use competitors instead of product
• Fixed risk assessment: reordered API call steps
• Added sellerLocation null fallback for Chinese seller analysis
• Added --keyword-match-type (fuzzy/phrase/exact) to prevent false matches
• Expanded confused fields table across all 3 interface types
• Fixed operations monitoring: threshold-based instead of trend-based
• Added realtime data structure notes to report/opportunity composites

✍️ New: Listing Optimization Module

• 8.1 Competitive Listing Analysis
• 8.2 Listing Copy Generation (title, bullets, description, search terms)
• 8.3 Listing Optimization Diagnosis (8-dimension scorecard)

🎯 6 Product Selection Mode Corrections

• beginner: added growth≥3%, listing≤1yr, 174 red ocean exclude keywords
• long-tail: added monthlySalesMax 300
• new-release: added FBA+FBM fulfillment
• fbm-friendly: added listingAge 180 days
• speculative: added listingAge 180 days
• top-bsr: changed to sub-category BSR (subBsrMax 1000)

🔧 New CLI Parameters

• --keyword-match-type, --bsr-min/max, --seller-count-min/max, --variant-count-max, --sub-bsr-max, --exclude-keywords

Full Changelog: v1.0.0...v1.1.0