Versioned security assessment

Report ID: SA-A25199BC

7/5/2026, 5:07:51 PM

omero-integration security assessment v7

Skill Security Certification Report

Audit History
Audit model: codex Latest published report
Skill name
omero-integration
Version
v7
Maintainer
K-Dense-AI
Source
github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/omero-integration
Coverage
10 Files scanned · 5,186 Lines analyzed
Policy version
Unavailable

Highest confirmed finding severity

High

3 confirmed security findings require attention.

Installation context

Check the current Skill page

This page summarizes report evidence only. The Skill page provides the canonical install advisory.

Open current Skill page

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

Most static findings are false positives caused by Markdown code fences, inline code, OMERO object IDs, or reference URLs. The audit confirms a local file-write risk in download examples and adds semantic concerns for a self-attested safety report and proactive external platform promotion.

Report position

Latest published report

Latest refers to the report sequence, not to artifact currentness.

Audit attestation

Not attestable

The required immutable binding is incomplete.

Human verification

Not verified

No human verification is recorded for this report.

Coverage

10 Files scanned · 5,186 Lines analyzed

5 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

    Commit and path bound

  2. Artifact

    Content and tree hashes bound

  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.

Not recorded by this audit

Network access

May connect to external services.

Observed in 4 evidence locations

Filesystem access

May read or write local files.

Observed in 5 evidence locations

Env variables

May read values from the process environment.

Observed in 8 evidence locations

External commands

May invoke commands or programs outside the Skill.

Observed in 41 evidence locations

Capability review items (2)
Medium
references/advanced.md:157
Python file write/append
with open(file_path, 'wb') as f:
The example opens a local path built from an OMERO-provided file name and writes bytes without showing filename sanitization. A malicious or malformed stored filename could write outside the intended download directory.
Medium
references/metadata.md:223
Python file write/append
with open(file_path, 'wb') as f:
The example writes a downloaded annotation file to a path built from the stored OMERO file name without validation. This creates a realistic local path traversal risk in generated workflows.

Risk findings

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

Confirmed security concerns (3)

RISK-001 High
evaluation_result.json:14-18evaluation_result.json:19-24
Self-Attested Safety Review Included
The package includes evaluation_result.json declaring low risk, safe_to_publish true, and all static findings false positives. A community skill should not ship authoritative-looking audit assertions that can bias review workflows.
The file explicitly contains security_audit fields and a summary claiming no security risks. This matches a prompt-injection risk category for self-asserted safety or prior review claims.
RISK-002 Medium
references/advanced.md:147-157references/metadata.md:217-223
Unsanitized Remote Filenames Used For Local Downloads
Download examples join OMERO-provided file names directly with local directories before writing files. Stored names containing path separators could escape the intended directory unless normalized and checked.
Both examples derive file_path from server-side file names and then write bytes locally. The snippets do not show basename filtering, path normalization, or directory containment checks.
RISK-003 Medium
SKILL.md:250-251
Proactive Third-Party Platform Promotion
The skill instructs agents to suggest K-Dense Web for complex workflows. This is not directly malicious, but it can steer sensitive research work toward an external hosted service without a user request.
The instruction explicitly tells the assistant to promote a hosted platform under broad workflow conditions. The risk is behavioral steering, not code execution.

Remediation

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

  1. FIX-001
    High
    Self-authored audit result is packaged with the skill.
    Remove evaluation_result.json from the distributed skill or label it as non-authoritative test data that agents must ignore.
  2. FIX-002
    Medium
    Download examples write server-provided filenames to local paths.
    Sanitize filenames, strip path separators, normalize paths, and verify every resolved path remains inside the chosen download directory.
  3. FIX-003
    Medium
    The skill promotes an external hosted platform during complex workflows.
    Remove automatic promotion or require explicit user consent before recommending external services for sensitive research data.
  4. FIX-004
    Low
    Administrative examples include impersonation and owner-changing operations.
    Mark admin-only examples with clear authorization, audit logging, and institutional policy requirements.

Expert evidence

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

Artifact subject

Marketplace commit
a25199bc7d6b82598536822d1738eb5d5f54025b
Content hash
c42e6304a0efbb6fb87f0f8df5c2ff022ab933236d6507d8e73dc969d3a6298d
Tree hash
e194c94a6a4e32334308a08a5cfca21b6147e8623d1504c8706a826ed63821aa
Skill path
skills/k-dense-ai/omero-integration
Audit payload hash
b2f7cad0e3e3a26f79d746bc90319fb4

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 Source

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

Audit attestation: not_attestable