Competitor Intel
Private-company intelligence: collectors, SQLite pipeline, Bun API, Svelte dashboard.
Competitor Intelligence
Private-company intelligence pipeline: multi-source ingest, funding rollups, trust-tiered claims, scored companies—SQLite SSOT; dashboard and Hermes read the same data.
Start here: docs/EXECUTION_CHECKLIST.md · Product: docs/PRODUCT_BRIEF.md · Ops: docs/PIPELINE.md · Standards: docs/ENGINEERING.md
Quick start
cd ~/Documents/Competitor-Intel
uv sync
export CI_DB_PATH="$PWD/data/competitor_intel.db"
make daily-prod # production cron: strict funding + dedup, no inline Grok
make grok-refresh # Hermes X batch (separate cron, ~5×/day)
make rollup-all # company / funding / job rollups when not inlined in daily
make claims-audit-strict
make health-check # SQLite checks
Hermes: hermes cron + docs/SCHEDULING.md; manual: uv run python integrations/hermes/call_intel.py daily-prod
CLI: make cli ARGS="status" or uv run python apps/cli/intel.py daily --export.
Toolchain
| Layer | Stack |
|---|---|
| Collectors / worker | Python 3.12+ via uv |
| Database | SQLite (data/competitor_intel.db, override CI_DB_PATH) |
| Hermes | External agent + integrations/hermes/ (required for Grok/X) |
Monorepo
| Path | Role |
|---|---|
packages/py-collectors |
Collectors |
packages/py-core |
DB, ingest, alerts |
apps/worker |
daily_intel.py, automation |
apps/cli |
intel.py, run_intel.py |
integrations/hermes/ |
Hermes HTTP/CLI shim |
Details: docs/ARCHITECTURE.md.
Documentation
docs/README.md — index
| Start here | |
|---|---|
| docs/PIPELINE.md | Daily pipeline, collectors, verification |
| docs/ENGINEERING.md | Naming, structure, no dead code |
| docs/ROADMAP_PRODUCTION.md | Full audit backlog (v2+) |
| docs/PIPELINE.md | Data flow and Makefile |
| docs/HANDBOOK.md | Schema and operations |
