Skip to content

CHANGELOG.md

Latest commit

 

History

History
42 lines (29 loc) · 3.02 KB

CHANGELOG.md

File metadata and controls

42 lines (29 loc) · 3.02 KB
Raw

Changelog

All notable changes to this project will be documented in this file. The format is based on Keep a Changelog.

[Unreleased]

1.0.0 - 2026-05-30

Initial public release of Rival — an open-source competitive intelligence dashboard powered by the Tabstack API.

Core features

  • Competitor scanner — multi-page orchestration across pricing, docs, changelog, careers, blog, reviews, profile, homepage, and GitHub page types; field-level merge preserves prior scan data when a page returns empty
  • Intelligence Briefs — AI-generated competitive summaries with threat level, positioning analysis, and 7-axis scoring
  • Positioning Matrix (/matrix) — interactive SVG plot with manual score overrides and per-competitor matrix: false exclusion
  • Intel Feed — dashboard view of recent changes across all competitors, sorted by threat level
  • Deep Dive (/[slug]/deep-dive) — streaming research mode (fast / balanced) with cited sources and pre-built prompt templates for messaging, developer sentiment, and strategic moves
  • Self-profile (/[slug]) — track your own company alongside competitors with a dedicated brief view
  • Demo mode (/demo) — public multi-surface scanner with rate limiting (3/day per IP) and concurrency lock; no account required
  • MCP server — 8 read-only tools (list_competitors, get_competitor, get_competitor_data, get_intelligence_brief, get_deep_dives, list_recent_intel, get_competitor_diff, search_intel) accessible via stdio or HTTP transport at /api/mcp
  • Experience logging (/insights) — per-call telemetry with fallback rates, missing field analytics, and extraction quality scores
  • Scheduled scans — daily scans via Netlify background functions with configurable cron schedule
  • Quiver integration — optional nightly push of competitor reports to Quiver research layer

Tabstack API modules

Full endpoint wrappers in lib/tabstack/ covering extract-json, extract-markdown, generate, automate, and research — each with mandatory experience logging, explicit effort settings, and fallback contracts. See the README for the full API reference.

Fixed

  • Deep Dive citations now correctly extracted from metadata.citedPages (the actual Tabstack API response shape — the top-level citations field the SDK documentation implies is never populated)
  • Dashboard lede no longer duplicates the lead story content
  • Self-profile Intelligence Brief section was blank on /[slug] — now renders correctly
  • Scanner preserves previous scan field values when a subsequent scan returns empty for that field

Security

  • .claude/settings.local.json gitignored to prevent credentials in Claude Code Bash allow-lists from being committed
  • Placeholder config uses RFC-reserved *.example.com domains so a fresh clone can never accidentally trigger real Tabstack API calls against unintended targets