Launched this week
XHawk 0.99
Transform Coding Sessions & Code into a System of Context
735 followers
Transform Coding Sessions & Code into a System of Context
735 followers
Turn your session history into a knowledge base. XHawk CLI automatically captures AI sessions during every git push, mapping the agent's reasoning directly to your commits. Don't just ship code. Capture the intent, audit the logic, and build a searchable, collaborative record of how your software actually gets built with coding agents. Our AI decodes your entire codebase, generating dynamic learning paths and docs for agents and humans.
Free
Launch Team / Built With
XHawk has been an excellent tool for revisiting codebases I haven’t explored in years. Even without documentation, it provides deep insights into the structure, core concepts, and underlying foundations of a repository. The generated Learning Path is highly organized, intuitive, and practical, making it easy to ramp up quickly and navigate complex systems with confidence. Unified Search further enhances this by offering clear visibility into how multiple repositories interact and work together. The challenges XHawk creates closely mirror what I would recommend to a newcomer to help them get familiar with the codebase.
Are there any plans to integrate XHawk with Claude or Codex so it can be used directly within the development workflow while building and iterating on code?
OS Ninja
Hey Product Hunters! 👋
I’m Puneet Singh, Co-founder of XHawk.
Today, we’re finally dropping something we’ve been obsessed with: System of Context. The AI-native knowledge index that ensures you and your coding agents never have to "start from zero" ever again. 🚀
Engineering knowledge shouldn't live in scattered coding sessions or outdated READMEs. We built XHawk to turn your code and AI interactions into a live, indexable source of truth. By automatically syncing every AI coding session and git commit into a real-time knowledge base, we’ve created the shared memory your engineering org has been missing.
Why "System of Context" is a must-have for AI-native teams?
⚡ Zero-Config, 60-Second Onboarding We hate bloat as much as you do. Run xh init and you’re live in less than a minute. Your sessions sync on every git commit, get indexed instantly, and are served via our MCP server. Getting full context into a fresh Claude session takes less than 60 seconds. Period.
🧠 AI-Native Memory Every session with Claude, Codex, or Gemini becomes a permanent asset of your engineering team. Now you don’t just capture the final code; you capture the reasoning, prompts, and intent behind it. It compounds every day.
🧩 Living Knowledge Graph Stop hunting for the "why" behind a change. Watch your code, docs, and decisions auto-link into a unified, searchable index that evolves every time you git push.
🛡️ Kill the Hallucinations AIs fail when they lack specific knowledge. XHawk feeds your agents "negative knowledge" (what not to do) and cold, hard facts from your repo’s real index to stop hallucinations before they start guessing. Think of it as "Context7" for your internal libraries and source code.
🚀 Documentation that actually works Stop writing docs that expire in an hour. XHawk builds context directly from the work itself—PRs, agent sessions, and code ASTs—keeping everything current in real-time.
🛤️ Beautiful Learning Paths No more manual onboarding. We perform deep research on your codebase to automatically generate structured learning paths. It maps out how your systems interact, so any developer (human or AI) can get up to speed in minutes, not days.
📄 The "AGENTS.md" Standard Stop re-teaching your stack. XHawk generates and maintains AGENTS.md and CLAUDE.md files for every relevant folder in your source code —living manifestos of your project’s patterns, "negative knowledge" (what NOT to do), and architectural guardrails. It’s the ultimate cheat sheet that makes every LLM session hit the ground running.
Stop re-explaining your codebase to your LLM 🛑
Since our private beta, we’ve seen teams reclaim massive amounts of dev time and millions of tokens by letting XHawk handle the "lore" of their codebase. We’re here to help you move from fragmented silos to high-velocity, AI-augmented workflows. Our mission is simple: make your SDLC agents stateful. No more forcing your agents to grep through source code and reconstruct your foundational architecture from scratch every single time they start a task.
I’m hanging out in the comments all day! I’d love to hear how you’re currently managing context for your coding, review, planning, testing and SRE agents.
We are so grateful to the Product Hunt community for your support. We’ll be here all day to answer your questions and hear your feedback!
Would you like us to add any specific features?
Hey Puneet, great concept and congratulations on the launch. It’s a much-needed tool for projects with lots of devs and complicated details. Do you have any features lined up to optimize token usage when connecting to MCPs, since over time the context might get bloated?
OS Ninja
@lokesh_motwani1 Our MCP server can be used to query for compact guidelines. We think it will save companies 20% tokens easily. Thanks for your feedback!
Congrats on the launch! The idea of capturing the reasoning behind code, not just the code itself, is something engineering teams have needed for a long time. Most knowledge walks out the door when a developer leaves or a context window resets.
The AGENTS.md standard is an interesting move. Would be curious to know how it handles projects where the architecture has evolved a lot over time, and the early decisions are basically archaeology at this point.
OS Ninja
@andreitudor14 We update the AGENTS.md files every night based on code changes. The key idea is that we should capture the patterns and anti-patterns from the codebase automatically. Every developer should not start from scratch. The biggest challenge is that LLMs are stateless. Engineering teams are successful because they are stateful. Our goal is to bring that "collective intelligence" into the software development lifecycle. It will be mostly driven by agents in the future so we are optimizing for that.
Netlify
Hey PH fam 👋
Big congrats to @puneet_singh25and the XHawk team on the launch!
This one resonates with me personally. I've watched Puneet obsess over this problem, and the more I dug in, the more I realized how real the pain is.
Here's something most dev teams haven't fully reckoned with:
Every time a new coding session starts, your AI forgets everything.
The reasoning behind that architectural decision. The three approaches you tried before landing on the right one. The "don't ever do this" lessons burned into your codebase. All of it vanishes. And someone, somewhere, has to reconstruct it from scratch. Every. Single. Time.
That's the problem XHawk is actually solving. Not just docs. Not just search. The intent behind the commits.
A context layer independent of any single LLM keeps your engineering knowledge portable and future-proof.
When the reasoning lives alongside the code permanently, something shifts. Your codebase stops being a mystery to new teammates and agents. Your best thinking compounds instead of evaporating.
That's a different kind of product. And I think it's the right one for where AI-native development is heading.
Go show Puneet some love and drop your questions below ⬇️
XHawk 0.99
Hello Product Hunters!
As developers, we’ve always felt the pain of scattered information. This has only become more prevalent as we move at high speed with AI agents. "Code is the truth!" many said but outdated documentation now causes more noise than clarity for our agents. We used to have very little reward for keeping docs up-to-date. Not anymore.
Humans are probabilistic. The more context and knowledge we gather, the fewer mistakes we make. The same is the case for Agents! That’s why we are building living documentation of the codebase.
We didn’t want to stop there!
Ever wondered why a developer wrote code in a certain way? We can finally answer these questions and avoid repeated mistakes. The code output is no longer a black box; we can finally capture the developer's thought process by capturing coding sessions and storing them alongside the codebase.
We didn’t want to stop there!
Have you ever wanted to search for a functionality but forgot the function name? Have you ever wished you could just ask: "How are the users created?" or "How is the JWT token created in the user flow?" This is where our search and agents come to the rescue. XHawk shows you relevant code snippets across all the repositories in your organization, providing quick answers inferred directly from the search results. We show you those code snippets alongside a summary of exactly what each snippet does.
We didn’t want to stop there!
For those developers who want to thoroughly learn a repository, we’ve got you covered. We have learning materials that target core functionalities without the fluff. But if you want to read it like a book? We have that, too!
As a developer, these were all the features I wished I had.
Thanks to coding agents, I can get quick solutions.
Thanks to XHawk tools, I finally have the context to design precise ones.
Happy Coding, everyone!
OS Ninja
@vinayak_suresh1 +100 on "without the fluff!"
I like that sessions sync on their own after each commit. You end up with a record of how features were actually built, not just the finished code.
OS Ninja
@samesh_lakhotia Yes, on git commit we sync the exact relevant script of the session.