index

tripleF (3F) — Technical Wiki

An agentic AI workbench, fully open source, powered by free models — run everything locally or on Ollama Cloud.

Document	Concern	Key Sections
0-introduction	Mission, feature set, roadmap, system overview	Why 3F, architecture diagram, technology matrix
0.1-quick-start	From clone to running stack	Prerequisites, environment, first run, verification
0.2-docker	Image design, compose topology, install model	Root Dockerfile targets, `deps` service, pnpm regimes, troubleshooting
0.3-documentation	Wiki conventions, badge automation	Writing docs, depbadge workflow, regenerating badges
1-server	Entry point, module graph, configuration stack	Request lifecycle, module table, error strategy
1.1-rest	Endpoint registry and contracts	Harness endpoints, queue/DLQ/storage/sysctl APIs, health
1.2-harness	Conversation engine internals	Step engine, intents, structured outputs, compaction, cancellation
1.3-bullmq	Job queues, workers, retries	Queue topology, processor pipeline, failure modes
1.4-socketio	Real-time streaming	Rooms, event contract, cancellation signal
1.5-data	PostgreSQL/Prisma, MinIO, KeyDB	Models, object layout, image preprocessing
2-dashboard	Areas and technology	Chat, SysCtl, DLQ, PProc — panel map
2.1-architecture	Vue 3 component and state architecture	Stores, composables, API layer, conventions
2.2-chat	The chat experience	Exchanges, reasoning areas, streaming UX, compaction
2.3-theming	Theme system	Theme catalogue, harmony model, tokens
3-ai-assisted-development	How this project is built	Context coding, ownership, quality gates
3.1-market-positioning	Where 3F competes	Proprietary landscape, local-first advantages, strategy