Dosu

Dosu is a knowledge infrastructure platform for AI agents that caches and shares information to reduce token usage and improve agent accuracy.

Dosu acts as a shared knowledge layer for AI coding agents, allowing information to be explored once and reused across different sessions. The platform connects to GitHub and GitLab repositories to monitor code changes, while syncing with documentation sources like Notion, Confluence, and Coda. It captures learnings from agent interactions and converts them into team-wide knowledge and reusable skills. This allows agents to access current context and best practices without requiring developers to change their existing workflows. The knowledge format is Markdown-based and compatible with the Open Knowledge Format, ensuring data remains portable and human-readable.

Dosu at a glance

Best for: Software engineer
Runs on: Web, GitHub, GitLab, Slack, Microsoft Teams
Pricing: Free plan for public repositories; Pro $192/billed upfront (4000 credits/month); Enterprise plans with custom pricing and audit logs.
Released: June 2023
Category: AI coding tool
Website: dosu.dev

Key features

Automatic documentation capture from agent sessions
Shared knowledge cache for token efficiency
Integration with GitHub, GitLab, Notion, Confluence, and Coda
Support for MCP (Model Context Protocol) servers
Self-documenting pull requests
On-demand documentation generation
Custom documentation templates

Use it for

Reducing token consumption for AI coding agents
Maintaining up-to-date technical documentation
Facilitating knowledge transfer within engineering teams
Integrating AI agents with internal enterprise knowledge sources

Alternatives to Dosu

AI Memory — Shared memory for coding agents that works across machines and team members.
CodeGraph — It eliminates file crawling, reducing AI agent tool calls by 88% and cutting costs.
Alook — Connects existing local coding agents to team communication channels.
CodeSight — Reduces AI token usage significantly and maps projects instantly.
Agent Beacon — Retains coding agent knowledge across tools so errors do not repeat.
Agent Skills — Encodes senior-engineer workflows and quality gates into 24 skills.
Kolo — Teams share one browser session to watch, type, and control AI agents together.
Rubi — It cuts AI token waste by keeping sessions short and mapping project knowledge.
Beeuni — Agents retain context and improve skills over time through a controlled review process.
AgentsKB — It supplies thousands of verified technical answers without hallucinations or stale data.
Factorai — Manages multiple independent AI coding agents in parallel from one interface.
TRAE — It features an autonomous coding agent that drives the entire development process.

All Dosu alternatives