Wow, what an awesome GitHub README. Informative and straight to the point.
I'm building an IDE for coding agents with a project management kanban board & plugin support as two of its flagship features. I think I can learn a lot from this project in its design as well as presentation.
Plugin sandboxing is an interesting tradeoff. I am planning full "video game coremod" style plugins with full access to the app and system, but it's interesting to see a WASM + sandbox approach.
As I use claude more and more I've started using git worktrees, one branch per worktree per PR, with possibly multiple agents working in each worktree at the same time on different aspects. And I manually instruct those agents. Like Emdash/Cursor/Zed. Sometimes I review code locally, sometimes agents push and I review in GitHub, no clear system yet. (jj seems promising, but Zed doesn't seem to support jj as well as git, so have delayed looking at it.)
But Paca is hinting in another direction where the agents are more in control of the branches/worktrees to use and are created by the agent? What tooling is used to support such flows? Would people use GitHub with Paca or is GitHub redundant as well.
This is pretty much my flow as well. Haven't gone beyond managing three work trees in parallel. It's nice being able to test locally against multiple work trees -- one is at 3000, then 3001, etc.
Gh issues works surprisingly well as an agent board. Labels for state, one issue per feature. The part i haven't figured out yet is how to know when the output is actually done vs just "looks done" to the agent.
I find well described but concise acceptance criteria does a good job of anchoring the llm to the correct output. Also have them take screenshots of any UI work and respond to the ticket with them as proof.
How much are worktrees benefitting you? If I can describe the work so clearly that it can be done in parallel, I find Claude can typically one shot so parallel work isn't needed.
You can take it a step further and strip out the frontend. Honestly. Nobody needs it and if you need any UI stuff it in the MCP.
This is what I did with this project https://github.com/crmkit/crmkit/ and to be honest the approach grows on me and fits well if you are a backend person.
I like the direction! But why stop with CRM functionality with very limited schema for tasks on the headless when OP has a lot more to offer there? Why stop at CRM plus Project Management then? Why not also schemas for ERP, HR, finance tools and all other business software? And we even bother with a custom made API when PostgreSQL and PostgREST could reduce the full thinking to the database schema?
I’m actually building an open-source Shopify for every vertical. The schema for each vertical is different and we are using Postgres. It comes with a built-in AI that you can ask to add new products, change prices, etc.
The next evolution for this is to allow users to use the AI to change the database schema itself. Like if someone is using our restaurant software and wants to start selling merch. I’d want them to be able to use the AI to change the database schema and add products and shipping from the e-commerce one.
I really do think that is the future of SaaS. You start from a base SaaS and the AI customizes it to your business over time.
I guess everyone uses 20% percent of Jira - just a different 20% ... [1]
We're using GitHub for everything here, but was using Jira as an email first helpdesk.
Was hoping this was that - but apparently not at all.
We almost went with libredesk - but it's a little too simple (no merging tickets?). We're giving FreeScout a go - looks like we might need the oauth2 plugin to work with o365 mail ...
> A lot of software developers are seduced by the old “80/20” rule. It seems to make a lot of sense: 80% of the people use 20% of the features. So you convince yourself that you only need to implement 20% of the features, and you can still sell 80% as many copies.
> Unfortunately, it’s never the same 20%. Everybody uses a different set of features.
Struggling to find a simple task management for work across different repos and some admin+marketing tasks. Become very messy when tasks are distributed among Hermes, Claude Code & myself.
Planned to building on beads to for non-repo based tasks, adding long term goal to keep driving agents busy, UI mainly sits on telegram group with topics. See whether Paca can solve the problem :)
I've been trying keeping an eye on open source issue trackers/project managament tools I can self-host -- with good cli/mcp capabilities. So quite happy to see this as I feel there isn't a lot! (currently also using gh issues) will check it out!
People who like Jira (or rather want; I doubt one ever “needs” this thing), and make decisions on its implementation and payment, and force it on others, are not the people who are shopping for alternatives. So who these alternatives are really for?
i always quite liked the flexibility of jira and the ability to logically connect tickets etc. I can see how it's perceived as this clumsy corporate tool, but i often whish gh issues had more of the features jira has.
This always irritates me, because it used to be just like that. First time we bought Jira, it was like a (small ) one-off charge on someone’s card, and included full source code to let you build it yourself.
I’m co-developing lots of projects with AI. Right now I have a hand-rolled backlog system that lives in each project’s git repo with a standard prompt on how to create, triage, and review backlog items.
This looks great for me. Better than what I have, smaller/cheaper/more AI focused than Jira.
I'm using GitHub issues and GitHub Projects with `gh` cli and I find it works well, though what I really like about this is your project level chat. I find myself having to come back to a project level session often. May give this a try, just hesitant to put it on something that's in-flight with already lots of stuff, will have to be a net new project.
Thanks! Glad you like the project-level chat. Starting with a fresh project is definitely the safest bet to see if the workflow fits you. If you ever do try it out, I'd love to get your feedback!
Feels like it's geared toward actually enabling the "dark factory", which is pretty difficult with enterprisey, seat based SaaS like GitHub and Jira. Will definitely check this out.
Awesome to see this. Like a few others here, I hand-rolled (well, Codex-rolled) something similar that works great for me. I keep going back and forth on open-sourcing it, but my hunch is people won't really adopt these kinds of things anyway.
Everyone ends up with a workflow shaped really tightly around how they work, and it's gotten so cheap to just build and evolve your own as the models and harnesses change that picking up someone else's stops making much sense.
I think we can consider this among the positive consequences of LLMs. Building software is cheaper, you don’t have anymore to adapt your company processes to the tools available in the market. If you don’t find what you’re looking for, you can build it and actually see if there’s a market interest for it.
Aren't we losing something there too though. I always respected a company with a product that had "things figured out" and pushed their product in conjunction with a way of working that was well researched and proven to be optimized.
I'm not convinced companies always need software tailored to their workflows, and could benefit from adopting worn-path workflows instead.
> I'm not convinced companies always need software tailored to their workflows, and could benefit from adopting worn-path workflows instead.
I’m dubious, because for an established company the question is whether the software adapts to the org, or if the org adapts to the software. It’s a lot harder to change the workflow of a whole company than to buy software that enables your current workflow. There’s months of retraining and figuring out where compliance goes in the new workflow, and things that get done wrong along the way because it’s new, and etc.
You need a pretty big efficiency win to offset the dead weight of time spent just changing workflows.
That makes sense when things are mostly stable and it makes little sense for most teams to work outside the norm.
Currently though we are in a world where things change every week, model capabilities, harnesses, pricing etc. Forcing a norm wont work, because there is no such norm.
I am fully convinced companies actually loose money because they have bunch of employees who waste time “bending reality” thinking they need custom workflow because “they are so specialized”.
Thanks for open-sourcing this! I built something similar for myself, but after few months it's so personalised that it's in no shape to be open-sourced.
That is exactly how this started! It's so easy for internal tools to become a mess of hyper-specific features.
I spent a lot of time trying to keep the core lean and moving the custom logic into the WASM plugin architecture precisely to avoid that trap. If you have any specific features from your internal tool that you found indispensable, I’d love to hear about them!
I think this is capturing the current need - solo vibe engineers that need structured task tracking. Since I pop between machines for various reasons, I tend to keep this info in the project itself, but an MCP server could go a long way. Tracking this project
Thanks! That's exactly why I built Paca. Traditional Jira feels way too slow when vibecoding—with this, we can just use the project chat to co-plan and assign tasks with AI in real-time.
Wow, what an awesome GitHub README. Informative and straight to the point.
I'm building an IDE for coding agents with a project management kanban board & plugin support as two of its flagship features. I think I can learn a lot from this project in its design as well as presentation.
Plugin sandboxing is an interesting tradeoff. I am planning full "video game coremod" style plugins with full access to the app and system, but it's interesting to see a WASM + sandbox approach.
What are people's workflows these days?
As I use claude more and more I've started using git worktrees, one branch per worktree per PR, with possibly multiple agents working in each worktree at the same time on different aspects. And I manually instruct those agents. Like Emdash/Cursor/Zed. Sometimes I review code locally, sometimes agents push and I review in GitHub, no clear system yet. (jj seems promising, but Zed doesn't seem to support jj as well as git, so have delayed looking at it.)
But Paca is hinting in another direction where the agents are more in control of the branches/worktrees to use and are created by the agent? What tooling is used to support such flows? Would people use GitHub with Paca or is GitHub redundant as well.
This is pretty much my flow as well. Haven't gone beyond managing three work trees in parallel. It's nice being able to test locally against multiple work trees -- one is at 3000, then 3001, etc.
Gh issues works surprisingly well as an agent board. Labels for state, one issue per feature. The part i haven't figured out yet is how to know when the output is actually done vs just "looks done" to the agent.
I find well described but concise acceptance criteria does a good job of anchoring the llm to the correct output. Also have them take screenshots of any UI work and respond to the ticket with them as proof.
How much are worktrees benefitting you? If I can describe the work so clearly that it can be done in parallel, I find Claude can typically one shot so parallel work isn't needed.
Not sure what you mean. Whether Claude one shots or not, it spends some minutes and in those minutes I can start another task in another worktree...
You can take it a step further and strip out the frontend. Honestly. Nobody needs it and if you need any UI stuff it in the MCP.
This is what I did with this project https://github.com/crmkit/crmkit/ and to be honest the approach grows on me and fits well if you are a backend person.
I like the direction! But why stop with CRM functionality with very limited schema for tasks on the headless when OP has a lot more to offer there? Why stop at CRM plus Project Management then? Why not also schemas for ERP, HR, finance tools and all other business software? And we even bother with a custom made API when PostgreSQL and PostgREST could reduce the full thinking to the database schema?
I’m actually building an open-source Shopify for every vertical. The schema for each vertical is different and we are using Postgres. It comes with a built-in AI that you can ask to add new products, change prices, etc.
The next evolution for this is to allow users to use the AI to change the database schema itself. Like if someone is using our restaurant software and wants to start selling merch. I’d want them to be able to use the AI to change the database schema and add products and shipping from the e-commerce one.
I really do think that is the future of SaaS. You start from a base SaaS and the AI customizes it to your business over time.
I need it, I want it. Why speaking out for me?
> You can take it a step further and strip out the frontend.
You can take it a step even further and strip out all the code too!
"Nobody needs it"
"Fits me well"
So maybe don't speak for everybody?
I guess everyone uses 20% percent of Jira - just a different 20% ... [1]
We're using GitHub for everything here, but was using Jira as an email first helpdesk.
Was hoping this was that - but apparently not at all.
We almost went with libredesk - but it's a little too simple (no merging tickets?). We're giving FreeScout a go - looks like we might need the oauth2 plugin to work with o365 mail ...
[1] https://www.joelonsoftware.com/2001/03/23/strategy-letter-iv...
> A lot of software developers are seduced by the old “80/20” rule. It seems to make a lot of sense: 80% of the people use 20% of the features. So you convince yourself that you only need to implement 20% of the features, and you can still sell 80% as many copies.
> Unfortunately, it’s never the same 20%. Everybody uses a different set of features.
-- Joel Splosky
Great project, will give it a try!
Struggling to find a simple task management for work across different repos and some admin+marketing tasks. Become very messy when tasks are distributed among Hermes, Claude Code & myself.
Planned to building on beads to for non-repo based tasks, adding long term goal to keep driving agents busy, UI mainly sits on telegram group with topics. See whether Paca can solve the problem :)
Thanks for having a security policy
https://github.com/Paca-AI/paca/security
However I'm getting a 404
https://github.com/Paca-AI/paca/security/advisories/new
(You need to enable private security advisories: https://docs.github.com/en/code-security/how-tos/report-and-..., really not sure why GitHub made it opt-in only)
Specifically on the AI side, how does it compare to beads?
I've been trying keeping an eye on open source issue trackers/project managament tools I can self-host -- with good cli/mcp capabilities. So quite happy to see this as I feel there isn't a lot! (currently also using gh issues) will check it out!
I can't believe you guys give this for free. I was considering buying "Linear", now I just saved 10$/month with this. Thank you so much
People who like Jira (or rather want; I doubt one ever “needs” this thing), and make decisions on its implementation and payment, and force it on others, are not the people who are shopping for alternatives. So who these alternatives are really for?
People in companies above 2-3 people have to use _some_ ticketing / task system.
I like Jira. Much more than Service Now
i always quite liked the flexibility of jira and the ability to logically connect tickets etc. I can see how it's perceived as this clumsy corporate tool, but i often whish gh issues had more of the features jira has.
This always irritates me, because it used to be just like that. First time we bought Jira, it was like a (small ) one-off charge on someone’s card, and included full source code to let you build it yourself.
I’m co-developing lots of projects with AI. Right now I have a hand-rolled backlog system that lives in each project’s git repo with a standard prompt on how to create, triage, and review backlog items.
This looks great for me. Better than what I have, smaller/cheaper/more AI focused than Jira.
There's an entirely new class of people doing development with AI.
Presumably some of them?
The people it's being forced on, maybe?
PM says "we're going to start using Jira", engineer says "how about we use this thing that looks similar but is not as terrible as Jira?"
Oh, what thing that would be? :)
Backlog.md the project: tasks live in your repo, atomic and race free
I'm using GitHub issues and GitHub Projects with `gh` cli and I find it works well, though what I really like about this is your project level chat. I find myself having to come back to a project level session often. May give this a try, just hesitant to put it on something that's in-flight with already lots of stuff, will have to be a net new project.
Thanks! Glad you like the project-level chat. Starting with a fresh project is definitely the safest bet to see if the workflow fits you. If you ever do try it out, I'd love to get your feedback!
Feels like it's geared toward actually enabling the "dark factory", which is pretty difficult with enterprisey, seat based SaaS like GitHub and Jira. Will definitely check this out.
When an agent and a human disagree on priority, who wins... is there an override, a queue or some kind of arbitration?
Could have called it AIpaca, since many fonts would make it look like Alpaca.
Awesome to see this. Like a few others here, I hand-rolled (well, Codex-rolled) something similar that works great for me. I keep going back and forth on open-sourcing it, but my hunch is people won't really adopt these kinds of things anyway.
Everyone ends up with a workflow shaped really tightly around how they work, and it's gotten so cheap to just build and evolve your own as the models and harnesses change that picking up someone else's stops making much sense.
I think we can consider this among the positive consequences of LLMs. Building software is cheaper, you don’t have anymore to adapt your company processes to the tools available in the market. If you don’t find what you’re looking for, you can build it and actually see if there’s a market interest for it.
Aren't we losing something there too though. I always respected a company with a product that had "things figured out" and pushed their product in conjunction with a way of working that was well researched and proven to be optimized.
I'm not convinced companies always need software tailored to their workflows, and could benefit from adopting worn-path workflows instead.
> I'm not convinced companies always need software tailored to their workflows, and could benefit from adopting worn-path workflows instead.
I’m dubious, because for an established company the question is whether the software adapts to the org, or if the org adapts to the software. It’s a lot harder to change the workflow of a whole company than to buy software that enables your current workflow. There’s months of retraining and figuring out where compliance goes in the new workflow, and things that get done wrong along the way because it’s new, and etc.
You need a pretty big efficiency win to offset the dead weight of time spent just changing workflows.
That makes sense when things are mostly stable and it makes little sense for most teams to work outside the norm.
Currently though we are in a world where things change every week, model capabilities, harnesses, pricing etc. Forcing a norm wont work, because there is no such norm.
I am fully convinced companies actually loose money because they have bunch of employees who waste time “bending reality” thinking they need custom workflow because “they are so specialized”.
Thanks for open-sourcing this! I built something similar for myself, but after few months it's so personalised that it's in no shape to be open-sourced.
That is exactly how this started! It's so easy for internal tools to become a mess of hyper-specific features.
I spent a lot of time trying to keep the core lean and moving the custom logic into the WASM plugin architecture precisely to avoid that trap. If you have any specific features from your internal tool that you found indispensable, I’d love to hear about them!
Was thinking about building something similar, thanks for sharing.
Glad to I'm not the only one thinking about moving away from Jira
This couldn’t have come at a better time!! This is exactly what I was going to build next now that my agent swarm is done.
We're using `twg` to give agents access to Jira et al via a CLI
https://www.atlassian.com/platform/teamwork-graph
Their skills abuse your context window and billing, you'll want to write your own for the 20% you use
I think this is capturing the current need - solo vibe engineers that need structured task tracking. Since I pop between machines for various reasons, I tend to keep this info in the project itself, but an MCP server could go a long way. Tracking this project
In my mind Jira is gone, glad to see others are thinking in the same direction.
Where does Jira really sit in a world eaten up by vibecoding?
Thanks! That's exactly why I built Paca. Traditional Jira feels way too slow when vibecoding—with this, we can just use the project chat to co-plan and assign tasks with AI in real-time.
great work , i love the ui and the smoothness is cherry on cake.