Colin, a context engine for keeping agent skills fresh
Details
- External ID
- 46769384
- Source
- HN
- Company
- —
- Product
- skills
- Website domain
- github.com
- Launched
- Jan. 26, 2026
- Cohort
- —
- Upvotes
- 7
- Upvotes percentile
- 0.3544137022397892
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:25 p.m.
- Updated at
- Sept. 7, 2026, 9:25 p.m.
Description
Hi HN — I'm the creator of FastMCP and wanted to share a new project we've open-sourced called Colin.I obviously love MCP, but I also use skills extremely heavily in my day-to-day work. Being exposed to both has made me very aware of a tension:- Anything with dynamic information, I ship over MCP. This takes work to set up and requires conversational boilerplate to refresh in every conversation.- Anything behavioral, I put in skills. They're lightweight, used automatically, and feel great. But I would never put dynamic information in a skill because keeping it up to date is a pain. And yet the idea of having an agent skill that includes relevant and timely data is so appealing!So: think of Colin as a skills-native templating engine. The goal is to compile context documents from dynamic sources (GitHub, Linear, Notion, MCP servers, etc.) in a way that makes them easy to automatically update. Colin traces dependencies to determine staleness and has intelligent caching to minimize the burden of recompiling. It's general purpose, but it turns out to be extremely useful for writing skills that contain dynamic knowledge and need to stay current.We've just released it. It's experimental and early, but it's a direction I'm actively exploring and would love feedback.Here's a post with more background, if interested: https://www.jlowin.dev/blog/colin
Enrichment
- Theme
- AI agent frameworks and developer tools
- Vertical
- Horizontal
- Function
- Agent / copilot
- Audience
- Developer
- AI stance
- AI-native
- Project type
- Commercial product
- Normalized one-liner
- context engine for agent skill management
- Manually corrected
- False
Could you build this?
Yes Colin is an open-source context/skill manager for MCP agents, which is fundamentally a stateful Python library and MCP server managing schemas and dynamic prompts.
Discussion
No comments on this launch.
Competitors
Other products that read as similar to this one — 314 launches clear the similarity bar, closest 8 shown.
Attention rank: #204 of 315 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 84 days after the earliest competitor.
- Skillscript · hn · 2026-07-12 · 18 upvotes · similarity 0.51
- An MCP server for Devops automation · hn · 2026-04-02 · 6 upvotes · similarity 0.45
- Keen Code · hn · 2026-08-10 · 6 upvotes · similarity 0.44
- Remy, an AI agent that compiles annotated Markdown into full-stack apps · hn · 2026-04-13 · 5 upvotes · similarity 0.44
- Cq · hn · 2026-03-23 · 225 upvotes · similarity 0.43
- Build agents via YAML with Prolog validation and 110 built-in tools · hn · 2026-01-23 · 11 upvotes · similarity 0.43
- Product analytics (and evals) for agent sessions on your MCP · hn · 2026-08-03 · 42 upvotes · similarity 0.43
- SubjectiveZero, an open-source agentic node editor for creative coding · hn · 2026-07-10 · 25 upvotes · similarity 0.43
Other launches for this product
- skills
- Claude Code skills that build complete Godot games
- Hivemind turns agent traces into skills and shares with your team
- skills
- adeskills
- skills
- skills
- Ratel, give agents unlimited tools and skills without context bloat
- Ingot, evidence-gated optimization and version control for agent skills
- Distributing AI agent skills via NPM
- Open-source skills for end-to-end video ad campaigns
- TLA+ Workbench skill for coding agents (compat. with Vercel skills CLI)
Same idea, different domain
Nobody's really built a agent / copilot tool for Agriculture yet.