← Back to Skills Marketplace
351
Downloads
0
Stars
0
Active Installs
4
Versions
Install in OpenClaw
/install r-stats
Description
82 statistical analysis methods in R — regression, survival, Bayesian, meta-analysis, causal inference, SEM, IRT, clinical trial design, and more. JSON spec...
Usage Guidance
This skill appears internally consistent for R-based statistical workflows, but it relies on executing included shell/R scripts. Before installing or running on real or sensitive datasets: 1) Inspect scripts/run-rstats.sh and all install-*.R and other scripts for network calls (curl/wget/git/R devtools::install_github), data exfiltration (HTTP POSTs, unusual endpoints), or commands that access files outside the declared output_dir. 2) Verify that R package installs come from CRAN/Bioconductor or trusted GitHub repos (no URL shorteners or personal servers). 3) Run the skill first in an isolated environment (container or sandbox) with non-sensitive example data. 4) If you want to allow autonomous invocation, prefer giving it only non-sensitive access or require manual approval for running scripts. If you’d like, provide the contents of scripts/run-rstats.sh and any install-*.R files and I will review them for unsafe operations.
Capability Analysis
Type: OpenClaw Skill
Name: r-stats
Version: 1.1.0
The skill bundle exhibits risky coding patterns and is missing core logic files, making a full security audit impossible. Specifically, 'tests/run_all_specs.sh' contains a command injection vulnerability where values from JSON files (like 'output_dir') are interpolated directly into an 'Rscript -e' command string without sanitization. This allows for arbitrary R and system code execution if a spec file is crafted maliciously. Furthermore, the primary execution scripts ('scripts/oc_rstats.R', 'scripts/doctor.R', and 'scripts/inspect_schema.R') are referenced by the bash wrappers but are absent from the bundle, which is a significant red flag for untrusted skill packages.
Capability Assessment
Purpose & Capability
Name/description (82 R methods) match the declared requirements: Rscript and bash are reasonable and sufficient for driving R analyses and package installs. The referenced R packages and method table in the repo are consistent with the stated capabilities.
Instruction Scope
SKILL.md gives a bounded workflow: confirm dataset, run a local schema script, build a JSON spec, run analyze, and read output files. Those steps are within scope for a stats skill. However the runtime flow depends on executing scripts (scripts/run-rstats.sh and many install-*.R files). The provided SKILL.md listing shows many example data/spec files but the script contents were not fully visible here. Any shell script executed by the agent can perform arbitrary actions (network I/O, exec other commands, read arbitrary files) so their exact behavior should be inspected before use.
Install Mechanism
No platform install spec (instruction-only) reduces supply-chain risk. The skill likely installs R packages via included install-*.R scripts as needed; this is expected for R-based tooling. Still, package installation is networked (CRAN/bioconductor/GitHub) and the install scripts should be checked for unusual download URLs, URL shorteners, or external servers.
Credentials
The skill declares no required environment variables, no primary credential, and no config path access. That is proportional: statistical analyses typically do not require secrets. There is no evidence the SKILL.md asks for unrelated credentials or reads environment variables beyond the declared binaries.
Persistence & Privilege
always:false and no claims to modify other skills or global agent configuration. The skill will run autonomously by default (normal for skills), but it does not request permanent platform-level privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install r-stats - After installation, invoke the skill by name or use
/r-stats - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.0
No user-facing changes. Version bump only.
v1.0.1
Fix: added runtime dependency declaration (Rscript, bash) to resolve OpenClaw security scan flag
v1.0.0
v1.0.0: Restructured per skill-creator spec. Method table moved to references/. Removed non-skill files. Added 10-language trigger keywords. Cleaner SKILL.md (<150 lines).
v0.7.0
82 reproducible statistical analysis methods in R. Covers regression, survival analysis, Bayesian inference, meta-analysis, causal inference, SEM, IRT, clinical trial design, and more. JSON spec driven with mandatory effect sizes, assumption checks, and standardized output.
Metadata
Frequently Asked Questions
What is R Stats?
82 statistical analysis methods in R — regression, survival, Bayesian, meta-analysis, causal inference, SEM, IRT, clinical trial design, and more. JSON spec... It is an AI Agent Skill for Claude Code / OpenClaw, with 351 downloads so far.
How do I install R Stats?
Run "/install r-stats" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is R Stats free?
Yes, R Stats is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does R Stats support?
R Stats is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created R Stats?
It is built and maintained by CuiweiG (@cuiweig); the current version is v1.1.0.
More Skills