Meeting recording
Drop in an mp3 or mp4. Cascade transcribes, identifies speakers, extracts user stories, and walks you through review before any code gets written.
cascade ingest standup.mp3cascade-agent 0.1.0a1 is live on PyPI
Meeting in.
Pull request out.
Open-source AI agent that takes a meeting recording, a tracker ticket, or a one-line prompt, and ships a tested pull request. Self-hosted. Your LLM key. Your code never leaves your org.
pip install cascade-agentThe pipeline
Eight stages. One terminal command.
Cascade orchestrates every stage from "I have a requirement" to "here's a tested PR." You stay in control at every gate. Review stories before code, review code before merge.
Ingest
audio to transcript
Extract
transcript to stories
Review
human gate
Plan
structured change
Code
files and tests
Test
pytest, vitest, go test
Commit
branch and push
PR
human review
Three on-ramps
Start from wherever the work lives.
Most AI tools assume requirements arrive as chat messages. Real teams discuss in standups, track in Jira, or just hand off a one-liner. Cascade accepts all three.
Tracker ticket
Point Cascade at a Jira, Linear, GitHub, GitLab, or Azure Boards ticket. It pulls the requirements and builds a PR with tests.
cascade ticket jira:PROJ-123One-line prompt
Type one line. Get a plan, code, tests, and a PR. Ideal for the small features that pile up on every team's backlog.
cascade prompt "Add /health endpoint"Plug into your stack
Bring the LLM, VCS, and tracker you already use.
Most AI tools force you onto their stack. Cascade meets you where you are. Five LLM providers, four VCS providers, five trackers, eight languages. Switch any of them with one config command.
LLMs
Anthropic OpenAI Google Gemini Claude Code Ollama
Version control
GitHub GitLab Bitbucket Azure DevOps
Issue trackers
GitHub Issues Jira Linear Azure Boards GitLab Issues
Languages
Python TypeScript JavaScript Go Rust Java Ruby C#
What makes it different
Built for real engineering teams.
01
Self-hosted by default
Cascade runs on your machine or your CI. Your code never leaves your org except to the LLM provider you explicitly choose.
02
Humans at every gate
Stories before code, code before merge. Cascade is built for trust, not for autonomy theatre.
03
Team memory
Conventions, decisions, and prior work in a bounded, structured grounding block. The AI gets context without prompt bloat.
04
Free forever, MIT
Use commercially, fork it, ship it inside your enterprise. No license keys. No usage caps. No telemetry.