Built on top of Jesse Vincent's superpowers. I tweaked the workflow to fit Rails development, added opinionated conventions, stricter reviews and deterministic checks. Happy to answer questions."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"My Rails harness for autonomous AI coding with Claude Code"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://rubyonai.com/my-harness-how-i-stopped-babysitting-ai-and-went-kitesurfing/"}},"_tags":["comment","author_marcinos","story_47453735"],"author":"marcinos","comment_text":"Author here. The workflow is packaged as an open source Claude Code plugin: https://github.com/marostr/superpowers-rails
Built on top of Jesse Vincent's superpowers. I tweaked the workflow to fit Rails development, added opinionated conventions, stricter reviews and deterministic checks. Happy to answer questions.","created_at":"2026-03-20T12:45:29Z","created_at_i":1774010729,"objectID":"47453767","parent_id":47453735,"story_id":47453735,"story_title":"My Rails harness for autonomous AI coding with Claude Code","story_url":"https://rubyonai.com/my-harness-how-i-stopped-babysitting-ai-and-went-kitesurfing/","updated_at":"2026-03-20T12:48:55Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dsifry"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"A few weeks ago I posted about GoodToGo https://news.ycombinator.com/item?id=46656759 - a tool that gives AI agents a deterministic answer to "is this PR ready to merge?" Several people asked about the larger orchestration system I mentioned. This is that system.
I got tired of being a project manager for Claude Code. It writes code fine, but shipping production code is seven or eight jobs \u2014 research, planning, design review, implementation, code review, security audit, PR creation, CI babysitting. I was doing all the coordination myself. The agent typed fast. I was still the bottleneck. What I really needed was an orchestrator of orchestrators - swarms of swarms of agents with deterministic quality checks.
So I built metaswarm. It breaks work into phases and assigns each to a specialist swarm orchestrator. It manages handoffs and uses BEADS for deterministic gates that persist across /compact, /clear, and even across sessions. Point it at a GitHub issue or brainstorm with it (it uses Superpowers to ask clarifying questions) and it creates epics, tasks, and dependencies, then runs the full pipeline to a merged PR - including outside code review like CodeRabbit, Greptile, and Bugbot.
The thing that surprised me most was the design review gate. Five agents \u2014 PM, Architect, Designer, Security, CTO \u2014 review every plan in parallel before a line of code gets written. All five must approve. Three rounds max, then it escalates to a human. I expected a rubber stamp. It catches real design problems, dependency issues, security gaps.
This weekend I pointed it at my backlog. 127 PRs merged. Every one hit 100% test coverage. No human wrote code, reviewed code, or clicked merge. OK, I guided it a bit, mostly helping with plans for some of the epics.
A few learnings:
Agent checklists are theater. Agents skipped coverage checks, misread thresholds, or decided they didn't apply. Prompts alone weren't enough. The fix was deterministic gates \u2014 BEADS, pre-push hooks, CI jobs all on top of the agent completion check. The gates block bad code whether or not the agent cooperates.
The agents are just markdown files. No custom runtime, no server, and while I built it on TypeScript, the agents are language-agnostic. You can read all of them, edit them, add your own.
It self-reflects too. After every merged PR, the system extracts patterns, gotchas, and decisions into a JSONL knowledge base. Agents only load entries relevant to the files they're touching. The more it ships, the fewer mistakes it makes. It learns as it goes.
metaswarm stands on two projects: https://github.com/steveyegge/beads by Steve Yegge (git-native task tracking and knowledge priming) and https://github.com/obra/superpowers by Jesse Vincent (disciplined agentic workflows \u2014 TDD, brainstorming, systematic debugging). Both were essential.
Background: I founded Technorati, Linuxcare, and Warmstart; tech exec at Lyft and Reddit. I built metaswarm because I needed autonomous agents that could ship to a production codebase with the same standards I'd hold a human team to.
$ cd my-project-name
$ npx metaswarm init
MIT licensed. IANAL. YMMV. Issues/PRs welcome!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: 127 PRs to Prod this wknd with 18 AI agents: metaswarm. MIT licensed"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/dsifry/metaswarm"}},"_tags":["story","author_dsifry","story_46864977","show_hn"],"author":"dsifry","children":[46866495,46866628,46866907,46867792,46868317,46870080],"created_at":"2026-02-03T01:18:39Z","created_at_i":1770081519,"num_comments":2,"objectID":"46864977","points":5,"story_id":46864977,"story_text":"A few weeks ago I posted about GoodToGo https://news.ycombinator.com/item?id=46656759 - a tool that gives AI agents a deterministic answer to "is this PR ready to merge?" Several people asked about the larger orchestration system I mentioned. This is that system.
I got tired of being a project manager for Claude Code. It writes code fine, but shipping production code is seven or eight jobs \u2014 research, planning, design review, implementation, code review, security audit, PR creation, CI babysitting. I was doing all the coordination myself. The agent typed fast. I was still the bottleneck. What I really needed was an orchestrator of orchestrators - swarms of swarms of agents with deterministic quality checks.
So I built metaswarm. It breaks work into phases and assigns each to a specialist swarm orchestrator. It manages handoffs and uses BEADS for deterministic gates that persist across /compact, /clear, and even across sessions. Point it at a GitHub issue or brainstorm with it (it uses Superpowers to ask clarifying questions) and it creates epics, tasks, and dependencies, then runs the full pipeline to a merged PR - including outside code review like CodeRabbit, Greptile, and Bugbot.
The thing that surprised me most was the design review gate. Five agents \u2014 PM, Architect, Designer, Security, CTO \u2014 review every plan in parallel before a line of code gets written. All five must approve. Three rounds max, then it escalates to a human. I expected a rubber stamp. It catches real design problems, dependency issues, security gaps.
This weekend I pointed it at my backlog. 127 PRs merged. Every one hit 100% test coverage. No human wrote code, reviewed code, or clicked merge. OK, I guided it a bit, mostly helping with plans for some of the epics.
A few learnings:
Agent checklists are theater. Agents skipped coverage checks, misread thresholds, or decided they didn't apply. Prompts alone weren't enough. The fix was deterministic gates \u2014 BEADS, pre-push hooks, CI jobs all on top of the agent completion check. The gates block bad code whether or not the agent cooperates.
The agents are just markdown files. No custom runtime, no server, and while I built it on TypeScript, the agents are language-agnostic. You can read all of them, edit them, add your own.
It self-reflects too. After every merged PR, the system extracts patterns, gotchas, and decisions into a JSONL knowledge base. Agents only load entries relevant to the files they're touching. The more it ships, the fewer mistakes it makes. It learns as it goes.
metaswarm stands on two projects: https://github.com/steveyegge/beads by Steve Yegge (git-native task tracking and knowledge priming) and https://github.com/obra/superpowers by Jesse Vincent (disciplined agentic workflows \u2014 TDD, brainstorming, systematic debugging). Both were essential.
Background: I founded Technorati, Linuxcare, and Warmstart; tech exec at Lyft and Reddit. I built metaswarm because I needed autonomous agents that could ship to a production codebase with the same standards I'd hold a human team to.
$ cd my-project-name
$ npx metaswarm init
MIT licensed. IANAL. YMMV. Issues/PRs welcome!","title":"Show HN: 127 PRs to Prod this wknd with 18 AI agents: metaswarm. MIT licensed","updated_at":"2026-03-05T23:29:47Z","url":"https://github.com/dsifry/metaswarm"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dsifry"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"A few weeks ago I posted about GoodToGo https://news.ycombinator.com/item?id=46656759 - a tool that gives AI agents a deterministic answer to "is this PR ready to merge?" Several people asked about the larger orchestration system I mentioned. This is that system.\nI got tired of being a project manager for Claude Code. It writes code fine, but shipping production code is seven or eight jobs \u2014 research, planning, design review, implementation, code review, security audit, PR creation, CI babysitting. I was doing all the coordination myself. The agent typed fast. I was still the bottleneck. What I really needed was an orchestrator of orchestrators - swarms of swarms of agents with deterministic quality checks.
So I built metaswarm. It breaks work into phases and assigns each to a specialist swarm orchestrator. It manages handoffs and uses BEADS for deterministic gates that persist across /compact, /clear, and even across sessions. Point it at a GitHub issue or brainstorm with it (it uses Superpowers to ask clarifying questions) and it creates epics, tasks, and dependencies, then runs the full pipeline to a merged PR - including outside code review like CodeRabbit, Greptile, and Bugbot.
The thing that surprised me most was the design review gate. Five agents \u2014 PM, Architect, Designer, Security, CTO \u2014 review every plan in parallel before a line of code gets written. All five must approve. Three rounds max, then it escalates to a human. I expected a rubber stamp. It catches real design problems, dependency issues, security gaps.
This weekend I pointed it at my backlog. 127 PRs merged. Every one hit 100% test coverage. No human wrote code, reviewed code, or clicked merge. OK, I guided it a bit, mostly helping with plans for some of the epics.
A few learnings:
Agent checklists are theater. Agents skipped coverage checks, misread thresholds, or decided they didn't apply. Prompts alone weren't enough. The fix was deterministic gates \u2014 BEADS, pre-push hooks, CI jobs all on top of the agent completion check. The gates block bad code whether or not the agent cooperates.
The agents are just markdown files. No custom runtime, no server, and while I built it on TypeScript, the agents are language-agnostic. You can read all of them, edit them, add your own.
It self-reflects too. After every merged PR, the system extracts patterns, gotchas, and decisions into a JSONL knowledge base. Agents only load entries relevant to the files they're touching. The more it ships, the fewer mistakes it makes. It learns as it goes.
metaswarm stands on two projects: https://github.com/steveyegge/beads by Steve Yegge (git-native task tracking and knowledge priming) and https://github.com/obra/superpowers by Jesse Vincent (disciplined agentic workflows \u2014 TDD, brainstorming, systematic debugging). Both were essential.
Background: I founded Technorati, Linuxcare, and Warmstart; tech exec at Lyft and Reddit. I built metaswarm because I needed autonomous agents that could ship to a production codebase with the same standards I'd hold a human team to.
$ cd my-project-name
$ npx metaswarm init
MIT licensed. IANAL. YMMV. Issues/PRs welcome!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Metaswarm: Production-ready agent swarms, MIT license"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://dsifry.github.io/metaswarm/"}},"_tags":["story","author_dsifry","story_46872673","show_hn"],"author":"dsifry","created_at":"2026-02-03T15:59:44Z","created_at_i":1770134384,"num_comments":0,"objectID":"46872673","points":1,"story_id":46872673,"story_text":"A few weeks ago I posted about GoodToGo https://news.ycombinator.com/item?id=46656759 - a tool that gives AI agents a deterministic answer to "is this PR ready to merge?" Several people asked about the larger orchestration system I mentioned. This is that system.\nI got tired of being a project manager for Claude Code. It writes code fine, but shipping production code is seven or eight jobs \u2014 research, planning, design review, implementation, code review, security audit, PR creation, CI babysitting. I was doing all the coordination myself. The agent typed fast. I was still the bottleneck. What I really needed was an orchestrator of orchestrators - swarms of swarms of agents with deterministic quality checks.
So I built metaswarm. It breaks work into phases and assigns each to a specialist swarm orchestrator. It manages handoffs and uses BEADS for deterministic gates that persist across /compact, /clear, and even across sessions. Point it at a GitHub issue or brainstorm with it (it uses Superpowers to ask clarifying questions) and it creates epics, tasks, and dependencies, then runs the full pipeline to a merged PR - including outside code review like CodeRabbit, Greptile, and Bugbot.
The thing that surprised me most was the design review gate. Five agents \u2014 PM, Architect, Designer, Security, CTO \u2014 review every plan in parallel before a line of code gets written. All five must approve. Three rounds max, then it escalates to a human. I expected a rubber stamp. It catches real design problems, dependency issues, security gaps.
This weekend I pointed it at my backlog. 127 PRs merged. Every one hit 100% test coverage. No human wrote code, reviewed code, or clicked merge. OK, I guided it a bit, mostly helping with plans for some of the epics.
A few learnings:
Agent checklists are theater. Agents skipped coverage checks, misread thresholds, or decided they didn't apply. Prompts alone weren't enough. The fix was deterministic gates \u2014 BEADS, pre-push hooks, CI jobs all on top of the agent completion check. The gates block bad code whether or not the agent cooperates.
The agents are just markdown files. No custom runtime, no server, and while I built it on TypeScript, the agents are language-agnostic. You can read all of them, edit them, add your own.
It self-reflects too. After every merged PR, the system extracts patterns, gotchas, and decisions into a JSONL knowledge base. Agents only load entries relevant to the files they're touching. The more it ships, the fewer mistakes it makes. It learns as it goes.
metaswarm stands on two projects: https://github.com/steveyegge/beads by Steve Yegge (git-native task tracking and knowledge priming) and https://github.com/obra/superpowers by Jesse Vincent (disciplined agentic workflows \u2014 TDD, brainstorming, systematic debugging). Both were essential.
Background: I founded Technorati, Linuxcare, and Warmstart; tech exec at Lyft and Reddit. I built metaswarm because I needed autonomous agents that could ship to a production codebase with the same standards I'd hold a human team to.
$ cd my-project-name
$ npx metaswarm init
MIT licensed. IANAL. YMMV. Issues/PRs welcome!","title":"Show HN: Metaswarm: Production-ready agent swarms, MIT license","updated_at":"2026-03-05T23:30:19Z","url":"https://dsifry.github.io/metaswarm/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"CharlesW"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"As a long time user, I'd recommend checking out https://github.com/obra/superpowers#the-basic-workflow. If that reasonates with you, try using it to develop a few features or capabilities.
It works very well for the way that I work (interactively and iteratively, not "one-shot"), and it helps me to better work in less time. Superpowers is one of the few skill/agent suites I use for all software development projects.
If you like building skill/agents, the posts at https://blog.fsck.com/ are a great resource for learning how to do well. The effectiveness of my project Axiom (a skill/agent suite for Apple OS developers) has benefited enormously from the knowledge that Superpowers' creator Jesse Vincent has been kind enough to share.
TLDR: You owe it to yourself to try it."},"story_title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["superpowers"],"value":"Superpowers 6"},"story_url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["superpowers"],"value":"https://blog.fsck.com/2026/06/15/Superpowers-6/"}},"_tags":["comment","author_CharlesW","story_48739459"],"author":"CharlesW","comment_text":"As a long time user, I'd recommend checking out https://github.com/obra/superpowers#the-basic-workflow. If that reasonates with you, try using it to develop a few features or capabilities.
It works very well for the way that I work (interactively and iteratively, not "one-shot"), and it helps me to better work in less time. Superpowers is one of the few skill/agent suites I use for all software development projects.
If you like building skill/agents, the posts at https://blog.fsck.com/ are a great resource for learning how to do well. The effectiveness of my project Axiom (a skill/agent suite for Apple OS developers) has benefited enormously from the knowledge that Superpowers' creator Jesse Vincent has been kind enough to share.
TLDR: You owe it to yourself to try it.","created_at":"2026-07-02T23:04:03Z","created_at_i":1783033443,"objectID":"48768514","parent_id":48768448,"story_id":48739459,"story_title":"Superpowers 6","story_url":"https://blog.fsck.com/2026/06/15/Superpowers-6/","updated_at":"2026-07-03T09:15:22Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Multiplayer"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"Superpowers is from Obra (Jesse Vincent), quintessential hacker and was a leader in the Perl community back in the day (still?).
https://github.com/obra/superpowers"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Why is my Claude experience so bad? What am I doing wrong?"}},"_tags":["comment","author_Multiplayer","story_47000206"],"author":"Multiplayer","comment_text":"Superpowers is from Obra (Jesse Vincent), quintessential hacker and was a leader in the Perl community back in the day (still?).
https://github.com/obra/superpowers","created_at":"2026-02-15T22:34:10Z","created_at_i":1771194850,"objectID":"47028405","parent_id":47028371,"story_id":47000206,"story_title":"Ask HN: Why is my Claude experience so bad? What am I doing wrong?","updated_at":"2026-03-05T23:33:21Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"groundy"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"Superpowers is an open-source agentic skills framework that transforms AI coding agents into disciplined software engineers by enforcing non-negotiable workflows: design before code, tests before features, and structured review between every task. Created by Jesse Vincent in October 2025 and accepted into the Anthropic marketplace in January 2026, it accumulated over 27,000 GitHub stars in its first three months\u2014roughly 9,000 per month."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"[dead]"}},"_tags":["comment","author_groundy","story_47304954"],"author":"groundy","comment_text":"Superpowers is an open-source agentic skills framework that transforms AI coding agents into disciplined software engineers by enforcing non-negotiable workflows: design before code, tests before features, and structured review between every task. Created by Jesse Vincent in October 2025 and accepted into the Anthropic marketplace in January 2026, it accumulated over 27,000 GitHub stars in its first three months\u2014roughly 9,000 per month.","created_at":"2026-03-09T04:43:05Z","created_at_i":1773031385,"objectID":"47304955","parent_id":47304954,"story_id":47304954,"story_title":"[dead]","updated_at":"2026-03-09T04:45:52Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"simonw"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["jesse","vincent","superpowers"],"value":"I know TDD provides better results for coding agents from 6+ months of experience working this, plus confirmation from conversations with other practitioners. TDD is the key methodology used by the popular superpowers set of Claude skills by Jesse Vincent, for example.
I'm not going to be trying to irrefutably prove everything I write about in the Agentic Engineering Patterns book - that would require a credible research team and peer-reviewed papers, and that's not a level of effort I'm willing to put into this."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Writing code is cheap now"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://simonwillison.net/guides/agentic-engineering-patterns/code-is-cheap/"}},"_tags":["comment","author_simonw","story_47125374"],"author":"simonw","children":[47158101],"comment_text":"I know TDD provides better results for coding agents from 6+ months of experience working this, plus confirmation from conversations with other practitioners. TDD is the key methodology used by the popular superpowers set of Claude skills by Jesse Vincent, for example.
I'm not going to be trying to irrefutably prove everything I write about in the Agentic Engineering Patterns book - that would require a credible research team and peer-reviewed papers, and that's not a level of effort I'm willing to put into this.","created_at":"2026-02-25T19:23:55Z","created_at_i":1772047435,"objectID":"47156475","parent_id":47153979,"story_id":47125374,"story_title":"Writing code is cheap now","story_url":"https://simonwillison.net/guides/agentic-engineering-patterns/code-is-cheap/","updated_at":"2026-03-05T23:37:20Z"}],"hitsPerPage":8,"nbHits":8,"nbPages":1,"page":0,"params":"query=jesse+vincent+superpowers&hitsPerPage=8&advancedSyntax=true&analyticsTags=backend","processingTimeMS":17,"processingTimingsMS":{"_request":{"roundTrip":18},"afterFetch":{"merge":{"mergeLoop":{"prepareNextHit":2,"total":2},"total":2},"total":2},"fetch":{"query":12,"scanning":1,"total":14},"total":17},"query":"jesse vincent superpowers","serverTimeMS":18}