v0.1.0

Rstudio Research Agent

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 6:08 AM.

Analysis

The skill matches its stated R research workflow purpose, but it can run local R code and install R packages when used.

GuidanceThis appears safe to install as an instruction-only R workflow helper. Before use, remember that running R scripts, RMarkdown, Quarto, or .Rprofile files executes local code, so use it with trusted projects and review any package installation commands.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Unexpected Code Execution

SeverityMediumConfidenceHighStatusNote

subskills/run-analysis/SKILL.md

Run specified R script ... Source `.Rprofile` if available ... `source("scripts/01_analysis.R")`

The skill is explicitly designed to execute local R project code and load project startup configuration. This is expected for the stated purpose, but R scripts and .Rprofile files can execute arbitrary code with the user's local permissions.

User impactRunning an untrusted R project could read or change local files, install software, or perform other actions allowed to the current user.

RecommendationUse this skill only with trusted R projects, inspect .Rprofile and RMarkdown/Quarto code chunks before execution, and confirm before running unfamiliar scripts.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceHighStatusNote

subskills/debug-env/SKILL.md

`install.packages("package_name")` ... `BiocManager::install("package_name")` ... `remotes::install_github("username/package")`

The skill may generate commands to install third-party R packages from CRAN, Bioconductor, or GitHub. This is normal for R dependency management, but package installation introduces external code into the user's environment.

User impactInstalling unknown or unpinned packages can bring untrusted code or unexpected dependency changes into an R environment.

RecommendationReview package names and sources before installation, prefer trusted repositories, and use renv.lock or other pinning practices for reproducibility.

Rogue Agents

SeverityLowConfidenceHighStatusNote

subskills/create-project/SKILL.md

Generate `.Rprofile` for automatic renv activation

The skill may create a project-level R startup file that persists after the initial task and runs when the project starts. The stated use is limited to renv activation and is disclosed.

User impactFuture R sessions in the project may automatically run startup code from .Rprofile.

RecommendationReview the generated .Rprofile and keep it limited to expected project setup such as renv activation.