Open Source · Apache 2.0

Intent in.
Software out.

Label a GitHub issue. AI agents triage, plan, implement, review, and merge it. You watch it happen and chat back to adjust.

Get Started View on GitHub
8+
Pipeline Stages
60+
Config Options
0
Human Intervention

Everything you need for autonomous development

From issue triage to merged PR. A complete pipeline of specialized AI agents working in concert.

🐙

Scalable Claude Workflow

Runs Claude CLI agents in Docker containers with git worktrees, slash commands, and hooks. Full isolation per agent with resource limits and network security.

🏗️

Repo Prep

Ensures tests, CI, linting, and other hygiene processes are in place. HydraFlow leverages your existing quality gates to guarantee accurate, high-quality output.

🔍

Autonomous Triage

Full TriagePhase with title/body quality checks, duplicate detection, and manifest analysis. Auto-detects languages, build systems, test frameworks, and CI before promoting.

🧠

Intelligent Planning

A read-only agent explores your codebase, understands architecture, and posts a detailed implementation plan as an issue comment.

🧪

TDD Implementation

Tests are written first. Quality gates enforce passing tests before creating PRs.

👁️

Automated Review

A dedicated review agent checks correctness, style, and test coverage, then submits formal GitHub PR reviews.

⚡

CI/CD Integration

Monitors CI checks in real-time. Automatically retries on transient failures and auto-merges approved PRs when all checks pass.

🤝

Human-in-the-Loop

When agents hit walls (ambiguous requirements, failing CI, hard merge conflicts) issues escalate to hydraflow-hitl for human guidance. You stay in control.

💾

Agent Memory

Agents learn from past runs. A persistent memory system with digest compaction extracts patterns and feeds them back into future sessions.

🔄

Crash Recovery

JSON-backed state persistence tracks every in-flight issue. If HydraFlow restarts, it picks up exactly where it left off. No lost work.

🔀

Merge Conflict Resolution

A dedicated MergeConflictResolver uses agent-assisted fixing to resolve conflicts automatically. Only hard conflicts escalate to HITL.

📊

Metrics & Analytics

15+ counters tracking cycle time, success rates, agent costs, and throughput. Full observability into your autonomous pipeline.

✅

Post-Merge Retrospective

After merge: acceptance criteria verification, retrospective analysis, epic progress checking, and structured learning extraction.

How it works

A label-driven pipeline that takes an issue from discovery to merged PR.

0
Issue

You + Claude

File an issue with context. Claude scans for duplicates.

1
Triage

Orchestrator

Evaluates readiness and promotes labeled issues

2
Plan

PlannerRunner

Explores codebase, posts implementation plan

3
Implement

AgentRunner

TDD implementation in isolated worktrees

4
Review

ReviewRunner

Code review, CI monitoring, fixes

5
Merged

Auto-merge

Approved PR merges when CI passes, with human UAT criteria attached

⚠️

Human-in-the-Loop

Escalates when agents need help

See it in action

Watch an issue flow through the entire pipeline, from triage to merged PR.

hydraflow - issue #42
[hydraflow] Starting pipeline...
[Triage] Found issue #42: "Add user avatar upload" ✓ Promoted
[Planner] Exploring codebase... analyzing 23 files
[Planner] Implementation plan posted to issue #42 ✓ Ready
[Implement] Creating worktree: agent-issue-42
[Implement] Writing tests... 4 test cases ✓ Pass
[Implement] Implementing feature... PR #87 created ✓ Pushed
[Review] Reviewing PR #87... code quality check
[Review] All checks passed. Approving. ✓ Approved
[Review] Auto-merging PR #87 into main ✓ Merged
[hydraflow] Issue #42 complete. Intent in → Software out.

The dashboard

Real-time visibility into every issue, agent, and pipeline stage.

Work Stream

Pipeline
HydraFlow Work Stream showing issues flowing through triage, plan, implement, and review stages

System Workers

System
HydraFlow System tab showing background workers: Retrospective, Review Insights, Pipeline Poller, Memory Manager, Metrics Munger, PR Unsticker

Human-in-the-Loop

HITL
HydraFlow HITL queue showing 50 items awaiting human review with approve, skip, retry, and close actions

HydraFlow builds HydraFlow

Every feature, bug fix, and improvement to HydraFlow is triaged, planned, implemented, and reviewed by its own agents.

From zero to building itself in 3 days using Vibe-to-Value™ methods.

Get started in 4 steps

From zero to autonomous development in under 5 minutes.

1

Clone & Install

# Add HydraFlow as a submodule
git submodule add https://github.com/T-rav/hyrda.git hydra

# Install git hooks (pre-commit, pre-push)
make setup
2

Prep Your Repo

# Create pipeline labels on your GitHub repo
make prep

# Ensure your repo has tests, CI, and linting in place
# HydraFlow leverages these quality gates for accurate output
make quality  # lint + typecheck + test
3

Configure

# Set up environment variables
cp hydra/.env.example hydra/.env
# Edit .env with your GitHub token and Anthropic API key
4

Run & Label

# Start HydraFlow (backend + dashboard)
make run
# Dashboard at http://localhost:5556

# In Claude Code, create an issue with a slash command
/gh-issue "make the text pink"