Skip to content
Back to skills

Candidate Search

ASecurity

Compare independent solution attempts in isolated forks when a reproducible check can choose among them and the task budget permits extra attempts.

  • 42 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 11, 2026
ai-agentsgo

Security analysis

A100/100

Scanned November 11, 2026

npx -y skills add acyclic-labs/sdk --skill candidate-search --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Candidate Search?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Candidate Search
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/acyclic-labs-candidate-search/badge)](https://www.skillsdirectory.com/skills/acyclic-labs-candidate-search)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: candidate-search
description: Compare independent solution attempts in isolated forks when a reproducible check can choose among them and the task budget permits extra attempts.
---

Define the judge before launching candidates: the same executable checks, inputs, and acceptance criteria for every attempt. Prefer task tests or an objective metric. A model's confidence or another model's preference is not a deterministic judge.

Give each fork the goal, constraints, starting state, and judge. Each candidate works independently and returns its branch, evidence, and remaining failures. Choose the number of attempts within the user's budget and the expected value of exploring alternatives.

Run the judge against the actual candidate changes. Compare only candidates that meet the required checks; use the task's stated objective to choose among them. Inspect and merge the chosen branch, then run the judge on the integrated result. If none passes, retain the failure evidence and revise the approach rather than declaring a winner.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.