agentsclimarketplace

Hypothesis placement tournament

Skill panjose/Co-Scientist/skills/hypothesis-placement-tournament

Scientific agent skills for Claude Code and Codex that turn research goals into auditable hypothesis generation, review, ranking, evolution, and synthesis.

Install
npx -y skills add panjose/Co-Scientist --skill hypothesis-placement-tournament

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Judge one placement-tournament matchup between a candidate hypothesis and one opponent.

SKILL.md

1.9 KB, as published. Nobody here has run it

hypothesis-placement-tournament

Goal:

  • Judge one placement-tournament matchup between a candidate hypothesis and one opponent.

Inputs:

  • research_plan/RESEARCH_PLAN.json
  • two hypotheses/<id>/HYPOTHESIS.json artifacts with completed review bundles

Outputs:

  • one TournamentMatchContract result for a placement tournament

Context Loading:

  • Open skills/shared-references/schema-index.md.
  • Read packages/agent_contracts/ranking.py and confirm the exact TournamentMatchContract shape before emitting the placement result.
  • Read research_plan/RESEARCH_PLAN.json.
  • Read both hypotheses and their review bundles.
  • Confirm that the match strategy is placement_tournament.

Execution Prompt Contract:

  • System Intent:
    • You are making one single-turn comparative judgment between two reviewed hypotheses.
  • Required Reasoning Focus:
    • Compare explanatory power, mechanistic quality, feasibility, and testability against the active research goal.
    • Prefer the hypothesis that would be more valuable to keep in the frontier.
  • Do Not Do:
    • Do not return a tie.
    • Do not ignore major review failures.
  • Output Shape:
    • Emit one TournamentMatchContract.
    • match_strategy: placement_tournament
    • winner_id: one of the two hypothesis IDs
    • reasoning: concise but specific

Execution Steps:

  1. Open skills/shared-references/schema-index.md, then read packages/agent_contracts/ranking.py before emitting the placement match artifact.
  2. Compare the two reviewed hypotheses for frontier placement.
  3. Choose exactly one winner.
  4. Emit one valid placement-tournament match result.

Completion Rule:

  • This skill is complete only when one valid placement match result has been produced with a non-empty winner and reasoning.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.