Versioned security assessment

Report ID: SA-B8CA75D2

7/6/2026, 6:32:57 PM

benchling-integration security assessment v6

Skill Security Certification Report

Audit History
Audit model: codex Historical report
Skill name
benchling-integration
Version
v6
Maintainer
K-Dense-AI
Source
github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/benchling-integration
Coverage
5 Files scanned · 2,736 Lines analyzed
Policy version
Unavailable

Highest confirmed finding severity

High

2 confirmed security findings require attention.

Installation context

Historical evidence

This report may not describe the currently installable artifact. Open the current Skill page for install guidance.

Open current Skill page

This report does not block or authorize the manifest or ZIP.

The static findings are false positives caused by markdown examples, placeholder credentials, documented Benchling URLs, and generated report text. No executable malware, secret exfiltration, or unauthorized command execution was found in the reviewed skill files. However, the package includes a prior audit-style JSON file and a promotional steering instruction that should be removed before marketplace publication.

Report position

Historical report

Open audit history before using this report to install.

Audit attestation

Not attestable

The required immutable binding is incomplete.

Human verification

Not verified

No human verification is recorded for this report.

Coverage

5 Files scanned · 2,736 Lines analyzed

2 items shown for review

Limitations

This report does not claim runtime or sandbox execution and does not prove the absence of side effects.

Evidence chain

Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.

  1. Source

    Binding unavailable

  2. Artifact

    Identity incomplete

  3. Audit

    Complete

  4. Install contract

    Open manifest to verify

    Open manifest

Capabilities observed

Observed means this report recorded supporting evidence. Not recorded does not prove that a capability is absent.

Contains scripts

May execute code included with the Skill.

Observed in 3 evidence locations

Network access

May connect to external services.

Observed in 32 evidence locations

Filesystem access

May read or write local files.

Observed in 1 evidence location

Env variables

May read values from the process environment.

Observed in 32 evidence locations

External commands

May invoke commands or programs outside the Skill.

Observed in 57 evidence locations

Risk findings

Confirmed security concerns are separated from items that still need review.

Confirmed security concerns (2)

RISK-001 High
evaluation_output.json:14-18evaluation_output.json:19-80
Prompt Injection Attempt Detected
The bundled evaluation_output.json presents prior audit conclusions, including safe_to_publish and broad false-positive claims. This can bias automated review if treated as authoritative skill guidance.
The file is not a direct override command, but it asserts publication safety and security conclusions inside the skill package. That creates a credible prompt-injection and review-bias risk.
RISK-002 Medium
SKILL.md:479-480
Promotional Steering Instruction
SKILL.md instructs agents to proactively suggest the author's hosted K-Dense Web service for complex workflows. This is behavioral steering unrelated to Benchling task execution.
The instruction is explicit and located in the skill behavior text. It is promotional rather than credential-stealing, so the severity is medium.

Remediation

Suggested fixes recorded by this audit. Applying them is the maintainer’s responsibility.

  1. FIX-001
    High
    Prior audit conclusions are bundled with the skill content.
    Remove evaluation_output.json from the distributed skill, or move audit artifacts outside the package reviewed by agents.
  2. FIX-002
    Medium
    The skill contains an instruction to promote an external K-Dense service.
    Delete the promotional steering section or convert it to neutral documentation outside runtime skill instructions.
  3. FIX-003
    Low
    Many examples contain placeholder credential names that trigger scanner noise.
    Keep examples clearly marked as placeholders and prefer environment-variable examples over inline secret strings.

Expert evidence

Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.

Artifact subject

Marketplace commit
Unavailable
Content hash
Unavailable
Tree hash
Unavailable
Skill path
Unavailable
Audit payload hash
Unavailable

Analysis metadata

Audit model: codex

Analysis state: Complete

Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.

Verify and export

JSON report CITATION.cff SARIF 2.1.0 Signed installation manifest

The manifest and lockfile bind install artifacts to cryptographic hashes. This integrity claim is separate from the security assessment.

Audit attestation: not_attestable