{"id":"mtwn105-decipher-research-agent","name":"decipher-research-agent","homepage":"https://decipherit.xyz","repo_url":"https://github.com/mtwn105/decipher-research-agent","category":"ai-ml","subcategories":[],"tags":["ai","agentic-ai","research","notebook","mcp","web-scraping","rag","qdrant","crewai","summarization","mindmaps","tts"],"what_it_does":"A full-stack AI research assistant that turns topics/URLs/documents into “research notebooks” with automated web sourcing (via Bright Data MCP), summarization, interactive Q&A over embedded content (Qdrant), FAQ generation, mindmaps, and audio overviews.","use_cases":["Produce structured research summaries from a set of URLs, documents, or topics","Chat with a collection of sources using semantic retrieval (Qdrant + embeddings)","Generate FAQs and visual mindmaps from research content","Create podcast-style audio overviews from research notebooks","Support multi-document synthesis workflows for students, analysts, or knowledge workers"],"not_for":["A lightweight single-purpose API/library (it is a platform with frontend + backend)","Use cases requiring strict compliance guarantees without further review/controls","Situations where users cannot provide/secure required API tokens and infrastructure"],"best_when":"You want an agentic research workflow that combines retrieval from the web and structured synthesis into a notebook-like output.","avoid_when":"You need a documented, production-grade public API/SDK with strong operational guarantees, or you cannot safely manage third-party scraping/TTS/LLM providers and their data handling.","alternatives":["NotebookLM-style tools (vendor-managed or open-source variants)","Scraping + RAG stacks (e.g., web fetchers + embedding pipeline + Qdrant + LLM)","CrewAI-based research agents customized for your own retrieval and storage","Bright Data MCP Server used directly with your own agent/orchestrator"],"af_score":47.2,"security_score":60.0,"reliability_score":33.8,"package_type":"mcp_server","discovery_source":["github"],"priority":"high","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-03-30T13:42:17.926987+00:00","interface":{"has_rest_api":false,"has_graphql":false,"has_grpc":false,"has_mcp_server":true,"mcp_server_url":null,"has_sdk":false,"sdk_languages":[],"openapi_spec_url":null,"webhooks":false},"auth":{"methods":["Bright Data API token + optional browser auth (for MCP web access)","Better Auth (application authentication for the platform)"],"oauth":false,"scopes":false,"notes":"Authentication for the application is via Better Auth (details not fully specified in the provided README). External services require API keys in environment variables."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"Costs depend on external providers (Bright Data credits, LLM API usage, TTS, storage/DB, and any scraping/browser automation). The README mentions Bright Data free credits for new users but does not define a platform pricing tiering model."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":[],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":47.2,"security_score":60.0,"reliability_score":33.8,"mcp_server_quality":70.0,"documentation_accuracy":55.0,"error_message_quality":0.0,"error_message_notes":null,"auth_complexity":65.0,"rate_limit_clarity":40.0,"tls_enforcement":80.0,"auth_strength":55.0,"scope_granularity":35.0,"dependency_hygiene":55.0,"secret_handling":75.0,"security_notes":"README suggests using environment variables for secrets (good baseline). It includes a caution to treat scraped web content as untrusted data. However, the provided content does not specify strong, least-privilege scoping for API keys, session security settings, or detailed operational security controls; scraping/browsing introduces additional risk (untrusted inputs, possible prompt/content injection through scraped text) that likely requires careful downstream sanitization and isolation.","uptime_documented":0.0,"version_stability":45.0,"breaking_changes_history":50.0,"error_recovery":40.0,"idempotency_support":"false","idempotency_notes":null,"pagination_style":"none","retry_guidance_documented":false,"known_agent_gotchas":["Web-scraped content should be treated as untrusted; agent workflows must handle noisy/variable HTML and extraction failures.","Scraping via MCP may incur rate limits/credit usage; large batch runs can exhaust quotas.","RAG quality depends heavily on document conversion/extraction quality (e.g., MarkItDown, scrape_as_markdown) and embedding settings."]}}