This report does not block or authorize the manifest or ZIP.
Static command-execution and weak-crypto alerts are mostly false positives from Markdown code fences and ordinary text. The skill is legitimate AIOps guidance, but copied examples include Kubernetes-changing automation, outbound API calls, credential-bearing Datadog headers, and user-provided webhook delivery, so publication should include an operational-risk warning.
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
1 Files scanned · 1,851 Lines analyzed
4 items shown for review
Limitations
This report does not claim runtime or sandbox execution and does not prove the absence of side effects.
01
Evidence chain
Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.
The automation engine examples construct kubectl scale, restart, patch, isolate, and failback commands from action targets and parameters, then execute them with asyncio.create_subprocess_exec. This is legitimate AIOps behavior, but copied code could change production infrastructure if used without authorization checks, input validation, and change controls.
The command construction and subprocess execution are directly visible, and the handlers modify Kubernetes resources. The risk is operational misuse rather than confirmed malicious intent.
The examples create HTTP sessions for monitoring systems and post notifications to a webhook URL supplied in action parameters. This is expected for AIOps integrations, but unvalidated endpoints can expose alert messages, service names, and operational metadata.
Network calls and user-supplied webhook posting are explicit in the examples. No evidence shows malicious exfiltration, but the data-flow risk is real if implemented without allowlists.
Credential Parameters Used for Datadog API Headers
The Datadog integration accepts API and application keys and places them into request headers. This is a normal integration pattern, but implementations should load credentials from a secrets manager and avoid logging or hardcoding them.
The code clearly handles API keys, but the keys are constructor parameters rather than embedded secrets. The finding is a handling concern, not evidence of credential theft.
03
Risk findings
Confirmed security concerns are separated from items that still need review.
Identifier-Based Reconnaissance Alerts Are Low Risk
The system-reconnaissance matches identify generated execution and healing IDs based on action, target, service name, and timestamp. These identifiers can reveal service names in logs, but they do not enumerate host or environment details.
The exact lines create IDs from operational names and timestamps. The privacy concern is plausible but limited because no host discovery or broad enumeration is shown.
04
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.
Static Analyzer Weak-Crypto Matches Are False Positives
The reported weak-cryptography locations point to prose, dataclass fields, feature extraction, dashboard descriptions, and closing text. I did not find evidence of MD5, SHA-1, or another weak cryptographic algorithm at those locations.
Targeted line review shows ordinary descriptive text and field names, not cryptographic calls. This strongly supports a false-positive classification.
Markdown Code Fence Command Alerts Are False Positives
Several Ruby backtick execution findings correspond to Markdown code fences that delimit Python examples. They are documentation formatting, not executable Ruby shell backticks.
The reviewed locations are Markdown delimiters around code blocks. They do not create shell execution behavior inside the skill file itself.