AI news · October 13, 2026
Dropstone CLI launches a self-learning coding agent for the terminal
Dropstone CLI launched on Product Hunt on October 12 with a free plan and paid plans reported from $5 to $150; it keeps corrections as lessons across coding sessions.
- Product Hunt launch points
- 87
- reported paid plan range
- $5-$150
- model choices named
- 3
Dropstone CLI is a terminal coding agent that reads a repository, edits across files, runs commands and tests, and checks its work before reporting completion. Its Product Hunt launch says a correction can become a lesson that carries into later sessions. The tool also lists Fast, Pro and Heavy model choices, plan-before-build behaviour, command permissions, custom skills, parallel subagents, MCP tools and GitHub issue to pull-request support. A free plan is available, with the maker describing paid plans from $5 to $150.
The product is aimed at a real weakness in coding-agent use: every new session can forget the local rules and past mistakes. Persistent lessons can help, but they can also preserve a wrong instruction or make a future change harder to understand. Start with a disposable repository and one repeatable issue. Compare compile success, test coverage, correction count and total spend with your current agent, and inspect what the tool stores before using it on a private codebase.
What you can do with it
Open the product page and run Dropstone on a disposable repository with tests that fail before the task and pass after it. Correct one mistake, repeat the same class of issue in a new session, and inspect the saved lesson. Keep command permissions narrow and do not connect a private repository until the memory behaviour is clear.
Our take
Dropstone has a strong commercial angle because it ties memory to a coding result instead of selling another chat window. The launch claims are maker-reported and the product is new, so persistent lessons need testing for stale or harmful advice. A small repeatability trial can show whether the memory saves time or creates new review work.
Source: Dropstone / Product Hunt — Made With Models writes the brief; the reporting is theirs.