Claude Skills

“With a skill claude will know when to use it without you specifically prompting it.”

“they're incredibly slow, expensive, unreliable, and even slow to create”

“The amount of fiddling needed to adjust is nearly equal to doing it yourself from scratch.”

“Using an AI skill for a process like that is slower than a script, less trustworthy”

“Skills don't really accomplish that well and are more unreliable when your target audience is a B2B customer”

“Skills do work, as they ground the agent with constrained context for the task it's performing”

“delivered via free, open-source Custom GPTs and Claude Skills”

“Skills are not ignored if you use a router in front of them, and they are actually called.”

“There is no reason to believe Co-work/Routines/Skills will survive 5 years from now.”

“Skills help budget tokens and stabilize around specific outcomes.”

“It is literally impossible because with skills, you may need to install some local CLI”

“I can’t imagine trying to make sure everyone in my company had the latest skill and CLI on their machine.”

“I love skills too because they can be out of the way until I decide to use them”

“It's not deterministic but it definitely runs each skill and it's raised a bunch of issues”

“skills are sort of MCP 2.0 fixing some of the mistakes”

“Claude Skills are starting to look like a structural advantage for Anthropic”

“skills are a big improvement compared to what we had before”

“I have like 50 different ones for personal life, work, dating, whatever”

“Best methods I’ve observed -progressive loading (claude skills) & symbolic search (serena mcp)”

“I think Skills is definitely an improvement over MCP.”

“I've actually built up a set of Claude Skills for creating and using tickets”

“Wouldn't it make perfect sense to wrap this in a Claude Skill instead of distributing as a separate package?”

“I think Skills are the killer feature because they are a very composable tool”

“Skills feel great for shaping how Claude works inside a repo”

“we've personally had such great results adopting it, especially for performing high quality large migrations”

“Skills are lazy loaded prompt engineering. They are simple, but powerful.”

“you can get very consistent charts because they're generated programmatically, but you can use the LLM for what it's good at”

“allowing an agent to write and execute code to access data from a 3rd party system is a security nightmare”

“Otherwise, you should use skills, especially if CLAUDE.md gets too long.”

“this an optimized mechanism for claude code to retrieve relevant information based on the skill's frontmatter”

“Skills are good for context management as everything that happens while executing the skill remains invisible to the parent context”

“it seems I have to specifically tell it to use the "X skill" to trigger it reliably”

“I haven't had much luck with skills being called appropriately.”

“with Skills the LLM can also inspect the tool, and modify it if it will help you”

“I find the sandboxed execution to be painfully inefficient.”

“skills for instance are a better approach for most use-cases”

“Claude Skills are just good documentation wrapped into Anthropic's API in a proprietary way that's designed to foster lock-in.”

“I love the idea of relying on markdowns and filesystem as universal API”

“Having simple Claude Skills as markdowns + a little bit of Ruby made this a lot easier than I expected.”

“Claude Skills just a few weeks ago was one of those ideas that's obvious in hindsight”