agentleFS
BagelHoleBagelHole/BagelSkills/skills/decision-tribunal/SKILL.md

decision-tribunal

Skill2 starsChanged 4 months agoMITView on GitHub

Use this skill when a consequential product, architecture, vendor, build-versus-buy, or strategy decision has multiple credible options and deserves adversarial examination. Steelman competing cases, test assumptions against evidence and failure scenarios, account for reversibility and delay, and issue a conditional recommendation with explicit dissent.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    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:

    Put consequential choices on trial so the final recommendation survives strong opposition rather than winning a shallow pros-and-cons …

    line 19
  2. Distinctive:

    If a required input is retrievable from available tools or artifacts, retrieve it rather than asking the user.

    line 40
  3. Distinctive:

    Write the decision in one sentence, deadline, owner, hard constraints, and what is explicitly outside scope.

    line 46
  4. Distinctive:

    Test hidden assumptions, tail risks, incentives, lock-in, migration, opportunity cost, and failure recovery.

    line 64
  5. Distinctive:

    Do not manufacture balance when evidence overwhelmingly favors one option.

    line 83
  6. Distinctive:

    Use comparable time horizons and cost definitions across options.

    line 84
  7. Distinctive:

    If a cheap experiment can answer the decision in time, recommend the experiment first.

    line 86

≈ “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

Decision Tribunal

Outcome

Put consequential choices on trial so the final recommendation survives strong opposition rather than winning a shallow pros-and-cons contest.

This skill changes the agent's process: it requires evidence, explicit claim strength, and a checkable completion boundary. Load references/playbook.md only when deeper prompts or rubrics are needed. Use assets/output-template.md when producing the final artifact.

When to use

  • The user asks to compare consequential options or challenge a preferred decision.
  • A team is deadlocked on architecture, vendor, build-versus-buy, or product strategy.
  • The cost of a wrong choice or delayed choice is substantial.

Do not use when

  • The choice is low-cost, reversible, and can be answered with a quick experiment.
  • One option violates a hard requirement and no real decision remains.

Required inputs

  • Decision, owner, deadline, and available options
  • Hard constraints and weighted decision criteria
  • Evidence, estimates, experiments, and stakeholder concerns

If a required input is retrievable from available tools or artifacts, retrieve it rather than asking the user. Ask only when a missing answer represents authority, preference, or risk that tools cannot resolve.

Workflow

1. Establish jurisdiction

Write the decision in one sentence, deadline, owner, hard constraints, and what is explicitly outside scope. Remove options that fail hard constraints with evidence.

Complete when: the decision is bounded and real options remain

2. Build the evidence ledger

Separate facts, estimates, assumptions, and preferences. Record source, freshness, confidence, and which options each item favors.

Complete when: claims are typed and confidence is visible

3. Appoint strong advocates

Construct the strongest case for every credible option using the same criteria. Do not give the preferred option better evidence or framing.

Complete when: each viable option has its strongest fair case

4. Cross-examine

Test hidden assumptions, tail risks, incentives, lock-in, migration, opportunity cost, and failure recovery. Run pre-mortems and identify evidence that would reverse each advocate’s position.

Complete when: decisive assumptions and tail risks are exposed

5. Account for time and reversibility

Classify one-way/two-way aspects, cost of delay, option-preserving moves, and the cheapest experiment that could collapse uncertainty before the deadline.

Complete when: the value of waiting or experimenting is explicit

6. Issue opinion and dissent

Recommend an option, conditions, confidence, immediate actions, and review trigger. Include the strongest dissent and specify when it should prevail.

Complete when: the decision owner knows what to do and what would change the answer

Operating rules

  • Hard constraints are pass/fail; preferences belong in weighted criteria.
  • Do not manufacture balance when evidence overwhelmingly favors one option.
  • Use comparable time horizons and cost definitions across options.
  • A recommendation without reversal conditions is overconfident.
  • If a cheap experiment can answer the decision in time, recommend the experiment first.

Output contract

The final result must contain:

  • Decision statement and constraints
  • Evidence ledger with confidence
  • Steelman case for each viable option
  • Cross-examination and pre-mortems
  • Reversibility, delay cost, and option value
  • Recommendation, dissent, conditions, and review trigger

Distinguish observed, inferred, assumed, and unverified claims. Link to files, commits, screenshots, commands, logs, or supplied sources when those artifacts support a consequential conclusion. Keep the executive result concise; move detail behind appendices when necessary.

Common pitfalls

  • Pros/cons lists with no weighting or evidence.
  • Strawman alternatives included only to validate a favorite.
  • False precision in scores built on weak estimates.
  • Ignoring the cost of waiting for perfect information.

Verification checklist

  • The requested outcome—not merely the procedure—has been satisfied.
  • Every consequential claim is supported, qualified, or explicitly unverified.
  • The workflow's completion criteria are met in order.
  • Safety boundaries and forbidden side effects were respected.
  • The output follows assets/output-template.md or contains equivalent information.
  • Remaining unknowns have an owner, next action, or stated consequence.

Resources

  • references/playbook.md — deeper domain prompts and evaluation guidance.
  • assets/output-template.md — reusable final-output structure.
  • evals/evals.json — output-quality cases.
  • evals/trigger-evals.json — positive and negative description-trigger cases.

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.