Skip to content

fcontext

Context continuity across AI agents & sessions.

fcontext - Context continuity across AI coding agents & sessions | Product Hunt
🔄 Cross-Agent, Cross-Session 👥 Team Collaboration
Switch between Copilot, Claude, Cursor, Trae — your AI never starts from zero. Export/import experience packs. Every team member's agent shares the same domain knowledge.
🛡️ Industrial-Grade Delivery 🔒 Offline & Secure
Structured context + requirements tracking + document indexing = consistent, auditable output. All data local in .fcontext/. No cloud. No API keys. No telemetry.

The Core Problem

AI coding agents are powerful — but they have three critical blind spots:

  1. Session amnesia — Every new conversation starts from zero
  2. Agent isolation — Switch agents, lose all context
  3. Team fragmentation — Each team member's AI works in a silo
graph TD subgraph "Without fcontext" U1["Developer"] --> A1["Copilot<br/>Session 1"] U1 --> A2["Claude<br/>Session 2"] U1 --> A3["Cursor<br/>Session 3"] A1 -.-x|"context lost"| A2 A2 -.-x|"context lost"| A3 end
graph TD subgraph "With fcontext" FC[".fcontext/<br/>shared context"] U2["Developer"] --> B1["Copilot"] U2 --> B2["Claude"] U2 --> B3["Cursor"] B1 <--> FC B2 <--> FC B3 <--> FC end

For Individuals

AI delivers results, but you deliver process and experience.

Your expertise — how you approach problems, what patterns you've learned, what pitfalls to avoid — is lost every time a session ends.

fcontext captures and persists that experience:

What you lose today What fcontext preserves
Debugging conclusions from yesterday _topics/debugging-auth-flow.md
Architecture decisions across sessions _README.md (AI-maintained)
Document analysis results _cache/ (indexed, reusable)
Project-specific patterns _experiences/ (exportable)

Your AI gets smarter over time

graph LR S1["Session 1"] -->|saves topics| FC[".fcontext/"] FC -->|loads context| S2["Session 2"] S2 -->|adds knowledge| FC FC -->|richer context| S3["Session 3"] S3 -->|even more| FC

Each session builds on the previous one. Your AI accumulates understanding instead of starting from scratch.

For Teams & Enterprises

No single agent has all the context to do the job. Real work is distributed.

In production environments, context is fragmented across people, documents, and conversations:

graph TD subgraph "Reality in Teams" D1["Requirements<br/>(PDF/DOCX)"] D2["Domain Knowledge<br/>(in people's heads)"] D3["Architecture<br/>(past conversations)"] D4["Conventions<br/>(tribal knowledge)"] end subgraph "fcontext unifies" FC["_cache/"] --- R["_requirements/"] R --- T["_topics/"] T --- E["_experiences/"] end D1 -->|"fcontext index"| FC D2 -->|"AI writes"| T D3 -->|"AI maintains"| T D4 -->|"fcontext export"| E

Key benefits for enterprises

Concern How fcontext addresses it
Onboarding New member imports experience pack → instant project knowledge
Consistency All agents read the same structured context → uniform output
Traceability Requirements tracked with evolution history → auditable decisions
Compliance All data stored locally, no cloud dependency → security-first
Knowledge retention Team expertise persists in _topics/ and _experiences/ → survives attrition

Team knowledge flow

graph LR TL["Team Lead"] -->|"fcontext export"| GR["Git Repo<br/>(experience pack)"] GR -->|"experience import"| D1["Dev 1's Agent"] GR -->|"experience import"| D2["Dev 2's Agent"] GR -->|"experience import"| D3["New Hire's Agent"] TL -->|"experience update"| GR

Quick Start

pip install fcontext

cd your-project
fcontext init
fcontext enable copilot    # or: claude, cursor, trae, opencode
fcontext index docs/

Your AI agent now reads project context automatically on every session.

Supported Agents

Agent Command
GitHub Copilot fcontext enable copilot
Claude Code fcontext enable claude
Cursor fcontext enable cursor
Trae fcontext enable trae
OpenCode fcontext enable opencode
OpenClaw fcontext enable openclaw