Back to skill

Security audit

Volcano Plot Script

Security checks for vulnerabilities and agentic risk

Overview

This volcano-plot skill is mostly purpose-aligned, but it needs review because it can generate unsafe R code from unescaped user input and can overwrite arbitrary writable output paths.

Review before installing or using in automated workflows. Use only trusted input arguments, avoid running exported R scripts generated from untrusted values, and restrict outputs to a controlled workspace until the R-string escaping and output-path validation are fixed. Pin dependencies before reproducible or production use.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (3)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/main.py:242
Finding

Executable R Code Injection Through Unescaped Command-Line Values

Content
View full analysis
% mutate(regulation = case_when( .data[[pvalue_col]] < pvalue_thresh & .data[[log2fc_col]] > log2fc_thresh ~ "up", .data[[pvalue_col]] < pvalue_thresh & .data[[log2fc_col]] < -log2fc_thresh ~ "down", TRUE ~ "ns" )) # Create plot p <- ggplot(data, aes(x=.data[[log2fc_col]], y=negLog10_pvalue, color=regulation)) + geom_point(alpha=0.6, size=1.5) + scale_color_manual(values=c("up"="{args.color_up}", "down"="{args.color_down}", "ns"="{args.color_ns}"), labels=c("up"="Upregulated", "down"="Downregulated", "ns"="Not significant")) + geom_hline(yintercept=-log10(pvalue_thresh), linetype="dashed", color="gray") + geom_vline(xintercept=c(-log2fc_thresh, log2fc_thresh), linetype="dashed", color="gray") + labs(x="{args.xlabel}", y="{args.ylabel}", title="{args.title}") + theme_minimal() + theme(legend.position="bottom") # Save plot ggsave("{output_path}", p, width=10, height=8, dpi={args.dpi}) print(paste("Plot saved to:", "{output_path}")) ''' r_output_path = output_path.with_suffix('.R') with open(r_output_path, 'w') as f: f.write(r_script) print(f"R script exported to: {r_output_path}") ``` ...[truncated 2336 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/main.py:328
Finding

Arbitrary User-Writable File Overwrite Through Unrestricted Output Paths

Content
View full analysis
{args.log2fc_thresh}, p < {args.pvalue_thresh})") print(f" Downregulated: {stats[1]} genes (log2FC < -{args.log2fc_thresh}, p < {args.pvalue_thresh})") print(f" Not significant: {stats[2]} genes") # Export R script if requested if args.export_r: export_r_script(args, output_path) # Export significant genes if requested if args.sig_genes_output: export_significant_genes(data, args.sig_genes_output, gene_col) ``` The associated export functions also open or write their destinations without containment or overwrite checks: ```python r_output_path = output_path.with_suffix('.R') with open(r_output_path, 'w') as f: f.write(r_script) ``` ```python sig_genes.to_csv(output_path, index=False) ``` ### Technical Analysis The `--output` and `--sig-genes-output` arguments are converted directly into filesystem destinations. The implementation does not: - Resolve paths against an approved workspace directory - Reject absolute paths - Reject `..` traversal - Detect symbolic-link targets - Refuse replacement of existing files - Require confirmation before overwriting a destination Python's `open(..., 'w')`, pandas `to_csv()`, and Matplotlib `savefig()` create or truncate destinations according to the permissions of the running account. The generated R path is derived from the supplied plot path with `with_suffix('.R')`, so it can also target an arbitrary user-writable location. This behavior conflicts with the security checklist in `SKILL.md`, which states that output directories sho ...[truncated 1388 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
requirements.txt:1
Finding

Unpinned Third-Party Dependencies Reduce Supply-Chain Integrity

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (21)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 20)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 33)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 34)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 40)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 49)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 60)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 68)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 69)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 70)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 112)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 231)May include surrounding context.

md
- Packaged executable path(s): `scripts/main.py`.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding

The skill advertises executable scripts and file I/O behavior but does not declare any explicit tool scope, permissions, or allowed-tools boundaries. In an agent environment, this can lead to over-broad execution authority, making it easier for the skill to read or write unintended files if the runtime grants default access.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description says the skill is triggered when a user 'needs visualization of gene expression data' or 'publication-ready figures for bioinformatics analysis,' which are broad phrases that could overlap with many general analysis or plotting requests beyond volcano plots. It does not provide clear boundaries or negative examples distinguishing when this skill should activate versus other visualization skills.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The guidance recommends using the skill for 'data analysis tasks that require explicit assumptions, bounded scope, and a reproducible output format' and for tasks needing a fallback path, which are generic qualities shared by many skills rather than specific triggers for volcano-plot generation. This ambiguity can cause unintended invocation because it fails to define scope-specific activation constraints.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The audit-ready example instructs users to pass clinical free-text to a tool that is documented to accept DEG file paths, creating an unsafe mismatch between declared and demonstrated inputs. This kind of prompt/input confusion can cause downstream scripts or agents to process unexpected sensitive text, mishandle validation, or normalize unsafe usage patterns that bypass the intended data model.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
96% confidence
Finding

The dependency manifest leaves pandas unpinned, so builds may resolve to different versions over time, including versions with security flaws or breaking changes. In a skill that generates plotting code, this mainly creates supply-chain and reproducibility risk rather than a direct code-execution path by itself.

Content

Scanner excerpt · requirements.txt (reported line 1)May include surrounding context.

text
pandas
matplotlib
seaborn
numpy

Unverifiable Dependency: pandas has 1 known advisory(ies) (CVE-2020-13091 (** DISPUTED ** pandas through 1.0.3 can unserialize and execute commands from an)), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
87% confidence
Finding

The manifest does not pin pandas, and pandas has at least one known advisory in some historical versions, so it is impossible to verify whether installs are affected. In this skill context, the risk is somewhat constrained because the dependency is for data handling and plotting, but an unbounded version range still creates avoidable supply-chain uncertainty.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
96% confidence
Finding

matplotlib is unpinned, which means the installed version can vary by environment or over time. This increases supply-chain exposure and undermines reproducibility, though in this plotting-focused skill the direct security impact is limited.

Content

Scanner excerpt · requirements.txt (reported line 2)May include surrounding context.

text
pandas
matplotlib
seaborn
numpy

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
96% confidence
Finding

An unpinned seaborn dependency allows uncontrolled version selection during installation. That can introduce vulnerable or incompatible releases and makes the environment less deterministic, even if the package is used only for visualization.

Content

Scanner excerpt · requirements.txt (reported line 3)May include surrounding context.

text
pandas
matplotlib
seaborn
numpy

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
97% confidence
Finding

numpy is unpinned, so dependency resolution may pull in different versions, including potentially vulnerable ones. Because numpy is a foundational package, version drift can also affect transitive behavior and reliability across the skill's environment.

Content

Scanner excerpt · requirements.txt (reported line 4)May include surrounding context.

text
pandas
matplotlib
seaborn
numpy

Unverifiable Dependency: numpy has 16 known advisory(ies) (CVE-2014-1859 (Numpy arbitrary file write via symlink attack); CVE-2021-41495 (NumPy NULL Pointer Dereference); CVE-2021-33430 (NumPy Buffer Overflow (Disputed)) +13 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
90% confidence
Finding

numpy has multiple historical advisories, but the requirements file does not specify a version, so the actual exposure cannot be determined. Given numpy's central role in scientific Python stacks, leaving it unpinned increases the chance of pulling an affected release, though this skill's volcano-plot use case does not inherently amplify exploitation likelihood.

Content

No source excerpt is available for this finding.