autoresearch plugins

uditgoenka/autoresearch

Plugin Claude Code

Bundles 5 hooks

Autonomous improvement engine. 14 commands: iterate, plan, debug, fix, security, ship, scenario, predict, learn, reason, probe, improve, evals, regression.

evo-hq/evo

Plugin Claude Code

Bundles 7 skills, 2 commands, 3 agents, 6 hooks · 990 tokens together

Structured experiment-driven code optimization using tree search and parallel subagents.

RUC-NLPIR/Arbor

Plugin Claude Code

Bundles 11 skills · 739 tokens together

Run Arbor's hypothesis-tree research workflow inside Claude Code using your own model — no API key. Bundles the Arbor Agent Skill suite and registers the keyless arbor mcp tools (Idea Tree, evaluation, worktrees, guarded merges, reports).

Human-Agent-Society/CORAL

Plugin Claude Code

Bundles 4 skills, 2 agents, 1 hook · 861 tokens together

Author and run CORAL tasks from your own harness. Teaches the coral CLI workflow (init/validate, start/status/resume/log/show) and checks coral is installed on session start.

gaasher/Agent-Loop-Skills

Plugin Claude Code

Bundles 25 skills · 3,870 tokens together

Drop-in agentic loop skills — autoresearch, literature & scientific writing, data analysis, code/SQL/prompt optimization, red-teaming, and power analysis. Each loop binds to your own task at invocation time.

CyrusNuevoDia/gepa-research

Plugin Claude Code

Bundles 2 skills · 136 tokens together

Research harness for optimizing code with the GEPA algorithm (LLM-driven genetic-Pareto search).

SeeleAI/Thoth

Plugin Claude Code

Bundles 1 skill, 11 commands · 292 tokens together

Dashboard-centric Agent Project OS for auditable AI execution, persistent project state, and human-visible workflow orchestration.

StepzeroLab/research-git

Plugin Claude Code

Bundles 3 skills, 3 agents · 435 tokens together

A memory system for the code you're exploring: capture each idea as a semantic Feature Capsule, recall it, and regenerate it onto today's codebase. Segmentation/regeneration run on natively-dispatched subagents (your subscription) — no pay-per-use API. MCP serves the graph read-only for sharing.

Zandereins/schliff

Plugin Claude Code

Bundles 18 skills, 1 hook · 1,148 tokens together

Deterministic quality scorer for AI instruction files (SKILL.md, CLAUDE.md, .cursorrules, AGENTS.md). 9 scorers incl. operationalcoverage, anti-gaming detection, autoresearch-style autonomous improvement.

proyecto26/autoresearch-ai-plugin

Plugin Claude Code

Bundles 2 skills, 1 command, 1 agent, 1 hook · 550 tokens together

Autonomous experiment loop for Claude Code. Run an endless optimize-measure-keep/discard cycle on any optimization target: LLM training loss, test speed, bundle size, build time, and more.

bayeslabs-rsi/Svatah

Plugin Claude Code

Bundles 6 skills, 3 hooks · 180 tokens together

Svatah structured experiment-driven code optimization using tree search and parallel subagents.

zircote-plugins/autoresearch

Plugin Claude Code

Bundles 1 skill · 173 tokens together

Autonomous skill improvement loop inspired by Karpathy's autoresearch pattern. Iteratively modifies skills, evaluates against fixed evals, and keeps or discards changes to converge on higher quality. Includes eval management and convergence reporting.

marcelopazzo/claude-code-autoresearch

Plugin Claude Code

Bundles 2 skills, 1 command · 113 tokens together

Autonomous experiment loop for Claude Code — run, measure, keep or discard. Ruby port of davebcn87/pi-autoresearch.