flamehaven-filesearch 1.3.1

FLAMEHAVEN FileSearch - Open source semantic document search with API authentication powered by Google Gemini

FLAMEHAVEN FileSearch v1.3.1

Self-hosted RAG search engine. Production-ready in 3 minutes.

Quick Start • Features • API Docs • Contributing

---

[>] Why FLAMEHAVEN?

[!] Fast	[#] Private	[$] Free
Production in 3 min	100% self-hosted	Generous free tier

---

[*] Features

Core

• Multi-format: PDF, DOCX, TXT, MD (up to 50MB)
• Semantic search: DSP v2.0 algorithm (zero ML deps, <1ms vectors)
• Search modes: Keyword, semantic, hybrid with typo correction
• Source attribution: Every answer links to source docs

v1.3.1 Enhancements

• Vector quantization: 75% memory reduction (int8)
• Gravitas-Pack: 90% metadata compression
• unittest suite: 19/19 tests, 0.33s runtime
• Zero heavy dependencies: No torch/transformers

Enterprise (v1.2.2)

• API key auth with permissions
• Rate limiting & audit logging
• Batch processing (1-100 queries)
• Admin dashboard

---

Problem → Solution

Before FLAMEHAVEN:

• Documents scattered across drives
• Keyword search too limited
• Privacy concerns with cloud services
• Complex RAG infrastructure

After FLAMEHAVEN:

• Local RAG in 3 minutes
• Intelligent semantic search
• Your data stays yours
• Single Docker command ??You don't want to upload your data to external services (Pinecone, Cloudflare, etc.) ??You need production-ready security without complex setup ??You want zero infrastructure costs for prototype phase

---

Solution: FLAMEHAVEN FileSearch

??Local RAG search engine in 5 minutes ??100% self-hosted (your data stays yours) ??Single Docker command deployment ??Free tier Google Gemini (up to 1500 queries/month) ??v1.2.2: Enterprise-grade authentication & multi-user support ??Batch search API for 1-100 queries per request ??Optional Redis for distributed caching across workers

---

[>] 3-Minute Quick Start

1. Docker (No Setup)

bash

Start with one command

docker run -d
-e GEMINI_API_KEY="your_gemini_api_key"
-p 8000:8000
-v $(pwd)/data:/app/data
flamehaven-filesearch:1.2.2

Available at http://localhost:8000 in 3 seconds

2. Your First Search (cURL)

---

[>] Quick Start

1. Docker (Fastest)

docker run -d \
  -e GEMINI_API_KEY="your_key" \
  -p 8000:8000 \
  flamehaven-filesearch:1.3.1

Access at http://localhost:8000

2. Python SDK

from flamehaven_filesearch import FlamehavenFileSearch, FileSearchConfig

config = FileSearchConfig(google_api_key="your_gemini_key")
search = FlamehavenFileSearch(config)

search.upload_file("document.pdf", "my_docs")
result = search.search("Summarize this", store="my_docs")
print(result['answer'])

3. REST API

# Generate API key
curl -X POST http://localhost:8000/api/admin/keys \
  -H "X-Admin-Key: your_admin_key" \
  -d '{"name":"prod","permissions":["upload","search"]}'

# Upload document
curl -X POST http://localhost:8000/api/upload/single \
  -H "Authorization: Bearer sk_live_..." \
  -F "file=@doc.pdf" -F "store=docs"

# Search
curl -X POST http://localhost:8000/api/search \
  -H "Authorization: Bearer sk_live_..." \
  -d '{"query":"main points?","store":"docs"}'

---

[&] Installation

# Core
pip install flamehaven-filesearch

# With API server
pip install flamehaven-filesearch[api]

# Development
pip install flamehaven-filesearch[all]

# Docker
docker build -t flamehaven-filesearch:1.3.1 .

---

[T] Configuration

Required:

export GEMINI_API_KEY="your_key"
export FLAMEHAVEN_ADMIN_KEY="secure_password"

Optional:

export HOST="0.0.0.0"
export PORT="8000"
export REDIS_HOST="localhost"  # For distributed caching

---

[=] Performance

Metric	v1.3.1	Notes
Vector generation	<1ms	DSP v2.0, zero ML deps
Memory footprint	75% reduced	int8 quantization
Metadata compression	90%	Gravitas-Pack
Test suite	0.33s	unittest (19/19 pass)

---

[#] Security

• API key SHA256 hashing
• Rate limiting (100/min default)
• Permission-based access control
• Audit logging with timestamps
• OWASP security headers

---

[W] Roadmap

v1.4.0 (Q1 2026)

• Multimodal search (images)
• HNSW vector index
• OAuth2/OIDC integration

v2.0.0 (Q2 2026)

• Multi-language support
• XLSX, PPTX, RTF formats
• WebSocket streaming

---

[L] Troubleshooting

401 Unauthorized:

• Verify FLAMEHAVEN_ADMIN_KEY is set
• Check Authorization: Bearer sk_live_... header

High memory usage:

• Enable Redis with maxmemory-policy allkeys-lru
• Monitor with /prometheus endpoint

Slow searches:

• Check cache hit rate in metrics
• Verify Gemini API latency

---

[B] Contributing

See CONTRIBUTING.md

Good first issues:

• Add XLSX support (2-3h)
• Dark mode dashboard (1-2h)
• Integration tests (2-3h)

---

Support

• CHANGELOG • Issues
• Security: [e-mail]

---

License

MIT License - see LICENSE

Last Updated: December 16, 2025 (v1.3.1)

[Your Documents] | v [File Upload Endpoint] ---> [File Parser] ---> [Store Manager] | | | +---- (REST API) --------+---- (SQLite DB)----+ | [Search Endpoint] ---> [Semantic Search] ---> [Gemini API] | | +--- (Cache) -------+ | [Prometheus Metrics] <--- [Audit Log]

---

Performance Metrics

Recent v1.2.2 benchmark (Docker on M1 Mac):

Health Check: 8ms Search (cache hit): 9ms Search (cache miss): 1250ms Batch Search (10 queries, parallel): 2500ms Upload (50MB file): 3200ms

---

License

FLAMEHAVEN FileSearch is released under the MIT License. See LICENSE file for details.

---

Acknowledgments

Built with:

• FastAPI - Modern Python web framework
• Google Gemini API - Semantic understanding
• SQLite - Lightweight database
• Redis (optional) - Distributed caching

---

Questions? Open an issue or email [e-mail]

Last Updated: December 09, 2025 (v1.2.2)