Hundao
589c5b06fe
- Auto-fixed 70 lint errors (import sorting, aliased errors, datetime.UTC)
- Fixed 85 remaining errors manually:
- E501: wrapped long lines in queen_profiles, catalog, routes_credentials
- F821: added missing TYPE_CHECKING imports for AgentHost, ToolRegistry,
HookContext, HookResult; added runtime imports where needed
- F811: removed duplicate method definitions in queen_lifecycle_tools
- F841/B007: removed unused variables in discovery.py
- W291: removed trailing whitespace in queen nodes
- E402: moved import to top of queen_memory_v2.py
- Fixed AgentRuntime -> AgentHost in example template type annotations
- Reformatted 343 files with ruff format
Competitive Intelligence Agent (Community)
Built by https://github.com/nafiyad
An autonomous agent that monitors competitor websites, news sources, and GitHub repositories to deliver structured digests with key insights and trend analysis.
Prerequisites
- Python 3.11+ with
uv - ANTHROPIC_API_KEY — set in your
.envor environment - GITHUB_TOKEN (optional) — for GitHub activity monitoring
Quick Start
Interactive Shell
cd examples/templates
uv run python -m competitive_intel_agent shell
CLI Run
# With inline JSON
uv run python -m competitive_intel_agent run \
--competitors '[{"name":"Acme","website":"https://acme.com","github":"acme-org"},{"name":"Beta Inc","website":"https://beta.io","github":null}]' \
--focus-areas "pricing,features,partnerships,hiring" \
--frequency weekly
# From a file
uv run python -m competitive_intel_agent run --competitors competitors.json
TUI Dashboard
uv run python -m competitive_intel_agent tui
Validate & Info
uv run python -m competitive_intel_agent validate
uv run python -m competitive_intel_agent info
Agent Graph
intake → web-scraper → news-search → github-monitor → aggregator → analysis → report
↑
(skipped if no competitors have GitHub)
| Node | Purpose | Tools | Client-Facing |
|---|---|---|---|
| intake | Collect competitor list & focus areas | — | ✅ |
| web-scraper | Scrape competitor websites | web_search, web_scrape | |
| news-search | Search news & press releases | web_search, web_scrape | |
| github-monitor | Track public GitHub activity | github_* | |
| aggregator | Merge, deduplicate, persist | save_data, load_data | |
| analysis | Extract insights & trends | load_data, save_data | |
| report | Generate HTML digest | save_data, serve_file | ✅ |
Input Format
{
"competitors": [
{"name": "CompetitorA", "website": "https://competitor-a.com", "github": "competitor-a"},
{"name": "CompetitorB", "website": "https://competitor-b.com", "github": null}
],
"focus_areas": ["pricing", "new_features", "hiring", "partnerships"],
"report_frequency": "weekly"
}
Output
The agent produces an HTML report saved to ~/.hive/agents/competitive_intel_agent/ with:
- 🔥 Key Highlights — most significant competitive moves
- 📊 Per-Competitor Tables — category, update, source, date
- 📈 30-Day Trends — patterns across competitors over time
Historical snapshots are stored for trend comparison on subsequent runs.