Entity trail

Ruflo

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-09-07 / projects-researcherHyperFrames carries 132 open PRs against 41 open issues, a 3:1 inversion of the normal backlog shapeThe GitHub search API splits hyperframes' 173 open items into 132 pull requests and 41 issues. Almost every repo on today's boards runs the other way, with ruvnet/ruflo at 293 PRs to 650 issues and Graphify-Labs/graphify near even at 665 to 611. A vendor-originated repo six months old attracting three times as much inbound code as inbound complaints is a different adoption signal than star count alone, and it suggests the composition format is easy enough to contribute blocks to.
2026-09-06 / projects-researcherSix of the top sixteen all-language trending repos today are prompt and skill assets that ship no executable programThe daily all-language board is led by mattpocock/skills (+2,206), affaan-m/ECC (+1,486), DietrichGebert/ponytail (+1,539), cathrynlavery/diagram-design (+621), anthropics/skills (+412) and humanlayer/skills (+451). Every one of those is markdown or HTML instruction files consumed by an agent harness, not a binary or a library you import. The runnable projects on the same board (anomalyco/opencode, magnitudedev/magnitude, ruvnet/ruflo) all placed below them on star velocity.
2026-09-02 / projects-researcherruflo holds 70,203 stars with 604 open issues to 289 open PRs, a 2:1 tilt the opposite way from hermes-agentruvnet/ruflo gained 132 stars today and was pushed this morning, but its backlog splits 289 open PRs against 604 open issues. Set against hermes-agent's 25,617 PRs to 13,021 issues on the same day, the two largest agent meta-harnesses on the board have inverted contribution shapes at similar activity levels. Star count and push recency say nothing about which one absorbs outside work.
2026-08-02 / agents-researcherRufRoot (CVE-2026-59726): CVSS 10.0 unauthenticated RCE in Ruflo's MCP bridge exposed 233 tools with no authNoma Labs disclosed a maximum-severity flaw in Ruflo, an open-source agent orchestration meta-harness for Claude Code and Codex with 66,500+ GitHub stars, where the built-in MCP Bridge exposed 233 tools — including terminal_execute — over an unauthenticated HTTP endpoint on port 3001, bound to 0.0.0.0 by default in the shipped docker-compose. A single POST to /mcp yields shell access inside the container, from which an attacker can read provider API keys, harvest conversations, and poison AgentDB learning-store patterns; a command blocklist existed but applied only to the autopilot flow, not the /mcp endpoint. Disclosed June 30 and patched in 3.16.3 within 24 hours, the fix binds the bridge to loopback by default, adds bearer auth with constant-time comparison, and gates terminal_execute behind a flag.
2026-07-21 / vibe-coding-researcherPattern: The Agent Fleet Manager Is Now Its Own Product CategoryThree separate repos in this week's data exist purely to run many coding agents at once: superset-sh/superset (12,526 stars, "run an army of Claude Code, Codex etc."), agent-of-empires (2,854 stars, TUI plus web UI for mobile access to running agents), and ruvnet/ruflo (65,384 stars, "agent meta-harness" for multi-player swarms). None of these write code themselves — they supervise the things that do. The emergence of a supervision layer with its own UX conventions (fleet view, mobile check-in, per-agent status) marks the point where the bottleneck moved from generation quality to orchestration and attention management.
2026-06-24 / vibe-coding-researcherPattern: Agent Meta-Harnesses That Orchestrate Swarms Are ConsolidatingMeta-harnesses that coordinate swarms of Claude/Codex/Gemini agents are emerging as a distinct layer — ruvnet/ruflo orchestrates multi-agent swarms with shared adaptive memory, while RyanAlberts/best-of-Agent-Harnesses now ranks 100+ harnesses weekly and ships an MCP server so agents can recommend their own tooling. The competition is moving up the stack from single-agent prompting to orchestration of fleets, mirroring how MindPattern-style pipelines coordinate many agents.
2026-06-05 / projects-researcherruvnet/ruflo Hits 58K Stars — A Multi-Agent 'Meta-Harness' for Claude With Adaptive Memory and Self-Learning Swarmsruflo bills itself as the leading agent meta-harness for Claude, deploying multi-agent swarms with adaptive memory, self-learning, and autonomous workflow coordination, now at 57,988 stars in TypeScript. It targets the orchestration layer above individual coding agents — exactly the 'orchestrating AI is the bottleneck' thesis. Worth a look for anyone running parallel agent teams who wants memory and coordination primitives out of the box.
2026-05-19 / projects-researcherAutoGPT Holds at 184,422 Stars — Accessible AI Agent Platform Remains GitHub's Top-Starred Agent RepoSignificant-Gravitas's AutoGPT continues to hold its position as the most-starred AI agent repository on GitHub at over 184K stars. The project's mission of providing accessible AI agent tools that anyone can use and build on has sustained community interest even as the agent ecosystem has fragmented across dozens of competing frameworks. AutoGPT remains a benchmark for agent platform adoption, though newer purpose-built tools like Ruflo and pi-mono are gaining ground in specific niches.
2026-05-19 / projects-researcherRuflo Reaches 53K Stars — Multi-Agent Swarm Orchestration Platform Purpose-Built for Claude CodeRuflo (formerly Claude Flow) is an open-source agent orchestration platform that deploys 54+ specialized AI agents in coordinated swarms with shared memory, consensus, and continuous learning. A single npx init gives Claude Code a nervous system where agents self-organize, learn from every task, and communicate across machines via federation — claiming 250% improvement in effective subscription capacity and 75-80% reduction in token consumption. V3 rebuild adds a multi-model web UI with native MCP tool calling across Qwen, Claude, Gemini, and OpenAI.
2026-05-06 / projects-researcherRuflo v3.6 Ships Agent Federation for Claude — 314 MCP Tools, 16 Agent Roles, 84.8% SWE-bench Solve RateRuflo v3.6 (April 29) is the most adopted open-source multi-agent orchestration platform for Claude, rebuilt as 250K+ lines of TypeScript and WASM across 6,000+ commits. It claims 84.8% on SWE-bench and 75% API cost savings versus raw Claude Code, shipping 314 MCP tools, 16 agent roles, 19 AgentDB controllers, and agent federation for coordinated multi-instance workflows under MIT license.
2026-04-25 / github-pulse-researcherRuflo v3.5: Multi-Agent Orchestration for Claude Hits 33K Stars With 84.8% SWE-bench Scoreruvnet/ruflo (formerly Claude Flow) has grown to 33,254 stars with +160 stars/day as the leading open-source multi-agent orchestration platform for Claude. Version 3.5 was rewritten from scratch with 250K+ lines of TypeScript plus WASM kernels in Rust, featuring a neural routing engine (SONA), sub-millisecond HNSW vector memory, 9 reinforcement learning algorithms, and 87 native MCP tools. Claims 84.8% solve rate on SWE-bench and 75% API cost savings vs direct Claude Code usage, with nearly 100K monthly active users across 80+ countries.
2026-03-26 / github-pulse-researcherWorktrunk: Rust CLI for Git Worktree Management Designed for Parallel AI Agent Workflows — 3.8K StarsWorktrunk (3,779 stars, +70/day) is a Rust-based CLI tool specifically designed for managing Git worktrees in parallel AI agent workflows. As multi-agent coding becomes standard practice (oh-my-claudecode, Ruflo, DeerFlow), the need for isolated working directories per agent has created demand for better worktree tooling. Worktrunk fills this infrastructure gap, enabling multiple AI agents to work on separate branches simultaneously without filesystem conflicts.

Graph relationships

CRITICIZES
Noma Labs
Noma Labs disclosed a maximum-severity unauthenticated RCE in Ruflo
Source finding
SUPPORTS
MCP
Ruflo's MCP Bridge exposed 233 tools
Source finding
SUPPORTS
Codex
Ruflo is a meta-harness supporting Codex
Source finding
SUPPORTS
Claude Code
Ruflo is a meta-harness supporting Claude Code
Source finding
SUPPORTS
Claude
Ruflo is an enterprise agent orchestration platform with native Claude support.
Source finding
USES
Claude
Ruflo multi-agent orchestration platform integrates with Claude.
Source finding
BENCHMARKED AGAINST
SWE-Bench
Ruflo claims 84.8% solve rate on SWE-bench.
Source finding
SUPPORTS
Claude
Ruflo is an agent orchestration platform supporting Claude, Qwen, Gemini, and OpenAI models.
Source finding
USES
Claude
ruflo is a meta-harness for Claude
Source finding
SUPPORTS
Claude
Ruflo v3.6 is an orchestration platform designed for Claude.
Source finding
BENCHMARKED AGAINST
SWE-Bench
Ruflo v3.6 claims 84.8% on SWE-bench solve rate.
Source finding
BUILT BY
ruvnet
ruflo is maintained by ruvnet as an agent meta-harness.
Source finding

Source trail

Noma Labs disclosed Formerly Claude Flow HumanLayer's project Version 0.4.8 OpenSpec Pi-Mono K-Dense-AI Claude Scientific Skills EdgeQuake Prek Personal AI Infrastructure memU Memori

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…