Spec Plan Build Review

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only software delivery workflow; it is coherent and disclosed, but users should be aware it includes commit, push, CI, release, and optional subagent review steps.

This skill appears safe as an instruction-only delivery workflow. Before installing or using it, be aware that the normal workflow may include reading repository context, running checks, committing, pushing, verifying CI, and preparing releases. Only use the shipping steps when you actually want remote repository or publication changes.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Risk analysis

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.

#
ASI02: Tool Misuse and Exploitation
Low
What this means

The agent may make persistent local and remote repository changes if the workflow reaches the shipping step.

Why it was flagged

These are high-impact repository and release actions. They are purpose-aligned with a delivery workflow, but they should be performed only when the user intends to ship.

Skill content
Commit with a specific message.\n   - Push and verify remote CI.\n   - Create release artifacts only after CI is green.
Recommendation

Confirm before allowing commits, pushes, tags, releases, or publication, especially for production or public repositories.

#
ASI07: Insecure Inter-Agent Communication
Info
What this means

Private code or task details could be shared with additional review agents in runtimes that support subagents.

Why it was flagged

The skill may use subagents for review when supported. This is aligned with the stated review purpose, but it can expand which agents see repository or task context.

Skill content
If the runtime supports real subagents and the task is large enough, fan out independent review passes.
Recommendation

Use subagent review only in trusted environments and avoid sharing sensitive code or secrets with agents that should not receive them.