Without Meldoc
› refactor session rotation
Most apps store sessions in Redis with a TTL of 24h. You can use a sliding window…
generic answer · no project context
Engineering context that doesn't rot in a wiki.
One place for engineering knowledge — written next to code, readable by your team and every AI tool.
1 member · 3 projects · 200 docs · core features including MCP
Your AI tools are working blind.
They don't have to.
Claude, Cursor, ChatGPT write code and plan features — but don't know your architecture, conventions, or how services connect.
Connect once via MCP. Every AI tool reads your engineering docs as structured, cross-repo context.
The tools you already pay for get dramatically better — because they finally have context.
With Meldoc MCP
› refactor session rotation
Your rotate_after is 30d with a 24h grace period. Move the rotation hook from
auth/service.go into…
cites your docs · matches your conventions
How it works
Step 01
Write next to code.
.meldoc.md in the repo, updated in the same PR as the code change.
Step 02
Sync happens on merge.
CI pushes docs automatically. Ready-made templates for GitHub, GitLab, Bitbucket.
Step 03
Everyone reads. AI reads too.
Web portal for the team.
MCP endpoint for Claude, Cursor, ChatGPT.
Repo .meldoc.md next to code
CI Validates on merge
For people
For AI
Three moments when it clicks
Onboarding: weeks, not months
New engineer asks any AI assistant instead of interrupting the team — it already has the full context.
On our own project, new people ship production code in week one.
Idea validation: a code-aware copilot for your system
Your assistant pulls architecture decisions and conventions automatically. Suggestions that fit your codebase, not generic advice.
Feature design: blockers surface early
PM's project chat knows the same constraints as the tech lead
Blockers at the brief — not after the sprint.
A second brain for your engineering team.
Meldoc builds a knowledge graph automatically — document hierarchy, [[wiki-links]], and semantic
similarity. Not just a visual. These connections are the structure underneath — what search navigates, what AI
follows when reading your docs through MCP.
auth-service
platform
tokens
session
rotation
refresh
oauth
login-flow
magic-link
mcp-server
portal-ui
rate-limit
db-schema
webhooks
Links Hierarchy Semantic Cross-project
What's included — all plans, all features
Only limits change. Everything below is in every plan, including Free.
MCP server
One endpoint. Claude Code, Cursor, ChatGPT read and write your docs as structured context.
CI/CD sync
GitHub, GitLab, Bitbucket. Drop the template, docs sync on every merge.
AI Ask
Question in, answer with citations out.
Knowledge graph
Hierarchy, wiki-links, semantic connections. Navigate visually.
Plus the basics
- Trust & access — Audit log, custom roles, webhooks, GitHub/Bitbucket sign-in, full markdown export.
- Workflow — Three-way merge, full CLI, public project publishing with llms.txt.
Simple, per-workspace pricing
Free for one. Flat per-workspace pricing for your whole team.
Free
Solo, internal docs
€0
Free forever
Included
- 1 member (solo)
- 3 private projects
- 200 documents
- 100 MB asset storage
- AI Ask (20/day)
- CLI + CI/CD workflow
- Basic search
- Single-project MCP (read-only)
- Magic Links
- 7 days history & audit log
Pro
Public docs & automation
€66 /mo
per workspace · billed annually
2 months free vs €79/mo monthly
Included everything in Free, plus:
- Unlimited members
- Unlimited projects & documents
- AI Search & Summarization
- AI Ask & Write (200/day)
- AI auto-links between docs
- Cross-project MCP + write access
- Bring your own AI key (BYOK)
- Webhooks
- 180 days history & audit log
- Email support
Need SSO/SAML, dedicated support, or custom export? Let's talk.
Contact UsFAQ
What is MCP?
Model Context Protocol — a standard that lets AI agents read external data. Meldoc serves your docs through MCP. Connect once, and Claude Code, Cursor, Claude Desktop, ChatGPT see your engineering context.
Do I need Git?
No. Write everything in the browser. The Git workflow — .meldoc.md files, CLI, CI sync — is there for teams that want docs next to code. Both paths lead to the same docs.
Can docs still go out of date?
Yes. If a developer changes code and doesn't touch the doc, nothing magic happens. The difference: with Meldoc, drift surfaces during PR review, in meldoc validate, in what AI reads. Drift becomes visible and fixable — not invisible and costly.
SAML SSO?
Not yet. GitHub and Bitbucket sign-in. SAML is on the roadmap.
Docs next to code. Context for AI. Readable by everyone.
Start free. Upgrade when you outgrow the limits.
1 member · 3 projects · 200 docs · core features including MCP