opensage-adk
Framework for building self-orchestrating, tool-heavy coding agents
All Skills in This Repository (32)
Pure Emerald Level Indicatorsopensage-adk
Orchestrate multi-agent sessions with sandboxed tools and Neo4j memory.
neo4j-query
Execute arbitrary Cypher queries against Neo4j graph databases to retrieve structured JSON records.
mmp
Multiplexes multiple MCP server connections via CLI across http, sse, stdio transports with persistent sessions.
create-new-tool
Scaffold bash tool directories with SKILL.md templates and executable scripts.
setup-matt-pocock-skills
Standardizes repository-level agent configs for GitHub, GitLab, and local markdown environments.
joern-query
Execute custom Joern queries against a code property graph for static analysis.
improve-codebase-architecture
Analyzes module structure and proposes deep redesigns for fragmented codebases.
triage
Manage GitHub issue lifecycle through state machine and role-based classification.
prototype
Build throwaway terminal applications to test state machines and UI variations.
diagnose
Diagnose software defects and performance regressions through structured hypothesis testing.
to-issues
Decompose project plans into granular vertical slice issues with dependencies and acceptance criteria.
zoom-out
Map modules and callers to provide architectural context for large codebases.
Frequently Asked Questions
FAQPage SchemaHow to install OpenSage-ADK?โผ
Run `npx skills add opensage-agent/opensage-adk --all -g -y` in your terminal to install all skills in this suite globally.
What does OpenSage-ADK do?โผ
It is a Python framework for building agents that dynamically create sub-agents, write their own tools, and manage structured memory during long tasks. It also bundles skills for static analysis, fuzzing, issue triage, and test-driven development.
What do I need to run OpenSage-ADK?โผ
You need Python 3.12 or later, the uv package manager, and Docker for sandboxed execution. Model access works through standard API keys like OPENAI_API_KEY or ANTHROPIC_API_KEY.
Can OpenSage agents run code safely?โผ
Yes. Every tool runs inside isolated Docker or Kubernetes sandboxes with configurable resource limits, so untrusted code execution stays contained.
Does OpenSage-ADK support multiple models?โผ
Yes. It uses LiteLLM, so you can mix models from OpenAI, Anthropic, Google, and others, and even run the same task across several models in parallel for comparison.
Related Repositories in Software Engineering
View All in Software Engineeringโopenclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core