performance-review

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 9:40 AM UTC
• Version: 1.9.19

Overview

This skill is a coherent performance-review aid, with expected source-code scanning and optional charting/install steps that users should invoke deliberately.

Install this only if you want an agent to inspect changed or specified source files for complexity hotspots. Invoke it explicitly, review any generated findings before acting, and pin or separately vet the optional kuva CLI before installing it.

A.I.G

Vulnerability Patterns

• Insecure Dependencies: Introduces malicious components through unsafe dependency sources
• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally

Findings (1)

T08 · Insecure Dependencies

Warning

• Location: modules/kuva-visualization.md:22

Finding

Unpinned Third-Party CLI Installation Creates a Supply-Chain Risk

Content

View full analysis

Vulnerability Details

File Location: modules/kuva-visualization.md:22
Vulnerability Type: T08: Insecure Dependencies
Risk Level: Medium

Complete Code Snippet:

bash

cargo install kuva --features cli

Technical Analysis

The installation command does not pin kuva to a reviewed version or require Cargo to honor a locked dependency graph. Consequently, the command installs whichever compatible release and transitive dependencies are available when it runs.

Cargo installations may compile packages and execute dependency build scripts. Therefore, following this instruction expands the trusted computing base to mutable upstream crate releases and their dependency chains. This is a supply-chain weakness rather than evidence that the current kuva package is malicious.

Attack Path

1. An attacker compromises the upstream crate, a maintainer account, or a transitive dependency.
2. A malicious release or dependency version is published to the package source used by Cargo.
3. A user or Agent follows the documented cargo install instruction after that release becomes available.
4. Cargo resolves and downloads the mutable package and dependency versions.
5. Malicious code may execute during compilation through build scripts or when the installed CLI is subsequently invoked.
6. The code runs with the permissions of the user who performed the installation or launched the CLI.

Impact Assessment

Successful exploitation could permit arbitrary code execution under the invoking user's account. Depending on that account's permissions, the attacker could access project files, benchmark inputs, environment variables, and other user-readable data, or modify files writable by that user.

No privilege-escalation mechanism, persistence mechanism, credential theft behavior, or known malicious upstream package is demonstrated by the audited files. The scope is limited to the authority of the user or ...[truncated 59 chars]

Remediation

View remediation

Remediation Suggestions

Pin the dependency to an exact version that has been reviewed, and require its published lockfile during installation:

bash

cargo install kuva --version '<reviewed-exact-version>' --locked --features cli

Additionally:

1. Record the approved version, expected publisher, and canonical source repository.
2. Review the selected release and its dependency graph before recommending installation.
3. Verify package provenance or checksums where the distribution workflow supports them.
4. Reassess security before updating the pinned version.
5. Prefer a project-managed, reproducible tool environment rather than an unrestricted user-global installation.
6. In CI, execute the visualization tool with minimal permissions and without unnecessary credentials or sensitive environment variables.

SkillSpector

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 92% confidence

Finding

Using the single-word trigger 'performance' makes the skill easy to invoke accidentally in unrelated conversations, reviews, or agent workflows. In agent ecosystems, overbroad activation can cause unintended file access, unnecessary analysis, workflow interference, or noisy outputs that displace more appropriate skills.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 90% confidence

Finding

The phrase 'code feels slow' is subjective and conversational, which raises the chance of accidental or contextually inappropriate activation. In a multi-skill agent environment, ambiguous triggers can lead to the wrong skill running on sensitive repositories or consuming resources without clear user intent.

Content

No source excerpt is available for this finding.

Vague Triggers

Low

• Category: Not specified by scanner
• Confidence: 84% confidence

Finding

This entry describes a situation rather than a concrete trigger phrase, making activation criteria unclear. A system matching on such language may invoke the skill too broadly during general merge discussions.

Content

No source excerpt is available for this finding.

Vague Triggers

Low

• Category: Not specified by scanner
• Confidence: 80% confidence

Finding

This trigger is phrased as a sentence fragment beginning with "or", which suggests descriptive prose rather than a precise activation condition. That reduces clarity about what exact user wording should invoke the skill.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low

• Category: Not specified by scanner
• Confidence: 87% confidence

Finding

The manifest describes a skill for detecting time and space complexity hotspots via AST/static analysis. The supporting module modules/kuva-visualization.md is documented as rendering benchmark data as charts, which is a separate reporting/visualization capability rather than an obvious requirement for AST-based hotspot detection.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.