Back to directory

No Negative Echo

An agent skill that makes coding agents regenerate commit messages, pull request titles, and other delivery text from the final, adopted state of the work, instead of carrying forward rejected drafts or intermediate discussion.

No Negative Echo

No Negative Echo is a skill, in the Agent Skills format, for AI coding agents. It addresses agents that revise their approach during a session but still leave traces of the rejected plan in the final commit message, code comment, pull request, or delivery note.

The skill has the agent regenerate delivery text from the adopted, verified final state rather than summarizing the conversation, and checks several delivery surfaces for leftover session residue. It gives the agent a short set of questions to decide what to keep: whether a reader of the final result needs the information, whether omitting it would be misleading or unsafe, and whether it describes a real, published change that the current surface needs to explain.

It can be installed as a skill with its own verification scripts, or its core rules can be copied into a project's AGENTS.md file for lighter, always-on enforcement. It is MIT licensed.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.