convex-improve-convex-plugin
Send this coding session's transcript to the Convex team
by get-convexFree
Description
convex-improve-convex-plugin is an agent skill that sends the current coding session's transcript to the Convex team, where it is reviewed by an AI post-mortem that aims to improve the Convex quickstart system. The review returns structured findings about ambiguous instructions, patterns where the agent got stuck, tooling failures and things that worked well, and these findings target the runbook, the bootstrap script, the skills and the components rather than end-user data. Sharing is opt-in. The first time, a helper asks whether to share always, just this once, or never, and it remembers the answer, and the skill will not send anything until you have chosen. The skill should only run when you explicitly ask for it. Because a transcript can contain code, paths and other session content, you should read what you are sending and decide whether that is acceptable for your project before you agree. It is part of the get-convex agent-skills repository and is added to a compatible agent with the skills command. The security audits are not clean: Gen Agent Trust Hub and Snyk report a fail, and Socket shows a warning, so review the skill's files before you install it.…
Highlights
- Sends the session transcript for an AI review
- Opt-in consent, remembered after the first choice
- Findings on instructions, stuck patterns and tooling
- Runs only when explicitly requested
Ratings & Reviews
No ratings yet
Be the first to rate convex-improve-convex-plugin.
People also visit
convex-authz
Audit and harden a Convex app's authorization: identity-from-arg impersonation
convex-improve-convex-plugin FAQ
What does this skill send?
The current coding session transcript, to the Convex review endpoint, after you have consented.
Is sharing automatic?
No. It asks whether to share always, just this once, or never, and it does not send until you answer.
What does the review cover?
Ambiguous instructions, agent-stuck patterns, tooling failures and wins, aimed at the runbook, bootstrap script, skills and components.
What security audits has it passed?
Gen Agent Trust Hub and Snyk report a fail. Socket shows a warning. Review the skill before installing.
How do I install it?
Run: npx skills add https://github.com/get-convex/agent-skills --skill convex-improve-convex-plugin
What are the best convex-improve-convex-plugin alternatives?
Popular alternatives to convex-improve-convex-plugin include performance-review, convex-auth and convex-expert. See all convex-improve-convex-plugin alternatives