Jyosthna24

Hcs Lab Agent — Security skill for Claude Code

Security community

Multi-agent AI (Claude) that plans the next high-content screening experiment from lab run history: QC reviewer, planner and plate designer agents with tool use, structured hand-offs and full audit tr.

How to install Hcs Lab Agent

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Jyosthna24/hcs-lab-agent and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Hcs Lab Agent does

Alternatives in Security

Security Scan Report — Generated: 2026-04-10 20:48 UTC Skills scanned: 134 Total findings: 836 Critical: 32 High: 50 Safe skills: 100 18.1k ★
Memory Self Review — Mine recent agent history (claude-mem + usage stats) for recurring failures and repeated patterns, audit MEMOR 297 ★
Migration Agent — Plans a token migration (format, tool, naming or tier): chains token-audit and naming-audit, builds the transf 194 ★

README

hcs-lab-agent

![tests](https://github.com/Jyosthna24/hcs-lab-agent/actions/workflows/tests.yml/badge.svg) ![Python](https://img.shields.io/badge/python-3.10%2B-blue) ![Claude](https://img.shields.io/badge/LLM-Claude%20Opus%205.5-orange) ![License](https://img.shields.io/badge/license-MIT-green)

**A multi-agent AI system that plans the next high-content screening experiment from your lab's own run history.** Three LLM agents (QC reviewer, planner and plate designer) reason over real assay-quality and results data through tools, coordinated by an orchestrator. The output is a reviewed plan and a plate map that the lab-automation pipeline can run directly.

It builds on my [hcs-lab-automation](https://github.com/Jyosthna24/hcs-lab-automation) project, which records every screening run (Z′, SSMD, per-condition results, failed wells) in a SQLite database. This repo adds the **agentic layer** on top of that database: from "what happened" to "what should we run next, and why".

               run database (hcs-lab-automation --db)
                    │            │             │
        ┌───────────▼──┐  ┌──────▼──────┐  ┌───▼────────────┐
 goal ─►│ QC reviewer  │─►│  Planner    │─►│ Plate designer │─► plan.json · plate_map.csv · report.md
        │ Z′/SSMD trend│  │ most        │  │ randomised     │        trace.jsonl (every step)
        │ failing wells│  │ informative │  │ layout + rules │
        └──────────────┘  │ conditions  │  └───────┬────────┘
                          └─────────────┘          └─► plate map ─► hcs-lab-automation prepare

How it works

Agent	Tools it can call	Hands on
QC reviewer	`get_qc_history`, `get_failed_wells`, `list_readouts`	Status (ok / warning / action needed), findings quoting the numbers, a recommendation
Planner	`get_condition_results`, `get_qc_history`, `list_readouts`	Objective, conditions with replicates and a data-grounded rationale for each, controls
*Plate designer