Graceful Boundaries

Security checks across static analysis, malware telemetry, and agentic risk

Overview

The skill appears to be a purpose-aligned HTTP conformance auditor, with expected live URL checks and optional local checker usage.

This looks safe to install for auditing Graceful Boundaries conformance. Be aware that it may make live HTTP requests to the URL you provide, and the bundle also contains implementation guidance that can edit a project if you ask for code changes.

Static analysis

No static analysis findings were reported for this release.

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 send HTTP requests to the URL being audited.

Why it was flagged

The skill performs live HTTP requests against a user-provided URL. This is expected for a conformance audit and is scoped to standard discovery paths, but users should ensure they are authorized to test the target.

Skill content
Fetch the limits discovery endpoint directly. Try both standard paths: GET <url>/api/limits; GET <url>/.well-known/limits
Recommendation

Use the skill only on services you own or have permission to assess, and avoid internal or sensitive URLs unless that is intentional.

#
ASI05: Unexpected Code Execution
Info
What this means

If used, the local checker runs JavaScript and makes HTTP requests to the audited service.

Why it was flagged

The skill documents optional execution of an included Node.js checker. This is purpose-aligned and not automatic, but it is local code execution that users should consciously choose.

Skill content
Optional accelerator: If the graceful-boundaries repo is cloned locally, the automated checker provides a structured report: node evals/check.js <url> --json
Recommendation

Run the checker only from a trusted copy of the repository/package and review results before taking action.

#
ASI02: Tool Misuse and Exploitation
Low
What this means

If the builder instructions are invoked, the agent may edit API/error-handling code in the current project.

Why it was flagged

The bundle includes a builder-oriented skill/instruction file that can change application code to implement the specification. This is related to the Graceful Boundaries purpose and appears user-directed, but it is broader than the primary audit description.

Skill content
this skill modifies source code in the current project
Recommendation

Use implementation mode only when you intend code changes, and review diffs and tests before deploying.