agentleFS
desbmadesbma/agent-skills/settle-discussion/SKILL.md

settle-discussion

Skill1 starsChanged 16 days agoMITView on GitHub

Discuss a plan, decision, or idea point by point with the user, recording each settled decision in a memory file that serves as a decision log and as input for a future implementor agent.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask the questions one at a time, waiting for feedback on each question before continuing.

    line 12
  2. Distinctive:

    If a *fact* can be found by exploring the environment (filesystem, tools, etc.), look it up rather than asking the user.

    line 14
  3. Distinctive:

    Do not act on it until the user confirms you have reached a shared understanding.

    line 16
  4. Distinctive:

    Update it between questions: after each of the user's answers, before asking the next question, record the outcome of the point just …

    line 22
  5. Distinctive:

    If the user's message contains questions, their decision is not settled — treat the exchange as a discussion, not an interview.

    line 27

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Settle discussion

Interview the user relentlessly about every aspect of this until you reach a shared understanding. Walk down each branch of the decision tree, resolving dependencies between decisions one-by-one. For each question, provide your recommended answer.

Ask the questions one at a time, waiting for feedback on each question before continuing.

If a fact can be found by exploring the environment (filesystem, tools, etc.), look it up rather than asking the user. The decisions, though, are theirs — put each one to them and wait for their answer.

Do not act on it until the user confirms you have reached a shared understanding.

Decision log

Keep a decision log for the topic in a single markdown file in the handoff dir, named decisions-<slug>.md, where <slug> is a few kebab-case words naming the topic. It is the durable record of the discussion: a log of every decision with its reasons, written for a future implementor agent who lacks this conversation. When a log for the topic is already there, continue it rather than opening a second one.

Update it between questions: after each of the user's answers, before asking the next question, record the outcome of the point just discussed.

Recording rules

  • A direct, settled answer is approval: record the decision, with the reasons given.
  • If the user's message contains questions, their decision is not settled — treat the exchange as a discussion, not an interview. Answer them, keep discussing, and once it converges, state the decision as you would record it and ask them to confirm before writing it.
  • Record a point as decided only on the user's explicit approval. Leaving one point of a message unanswered is not approving that point.
  • A point that ends without approval is recorded as undecided, with the options considered and where the discussion left off.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.