📦

Audit History

incident-runbook-templates - 8 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v8 LatestJul 7, 2026, 07:31 AM No confirmed findings0No capability change
v7 Jul 7, 2026, 07:31 AM No confirmed findings0No capability change
v6 Jul 1, 2026, 01:14 AM No confirmed findings1No capability change
v5 Jul 1, 2026, 01:14 AM No confirmed findings1No capability change
v4 Jan 17, 2026, 09:28 AM No confirmed findings0No capability change
v3 Jan 17, 2026, 09:28 AM No confirmed findings0External commandsFilesystem access
v2 Jan 4, 2026, 04:31 PM No confirmed findings1No capability change
v1 Jan 4, 2026, 04:31 PM No confirmed findings1Baseline

Jul 1, 2026, 01:14 AM

Static analysis flagged many command, URL, IP address, filesystem, and weak-crypto patterns in SKILL.md. Review found these are Markdown runbook examples and false positives, not executable skill code or malicious behavior. The skill is safe to publish with a low-risk warning because operators may copy impactful production commands.

1
Files scanned
384
Lines analyzed
4
Review items
2
False positives ignored
Capability review items (1)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Low
Operational Command Examples Require Review
Verdict: TRUE_POSITIVE as content risk, not executable malware. SKILL.md contains Kubernetes, curl, psql, and shell command examples for incident response. They are inside Markdown templates, so the skill does not execute them, but users could cause production changes if they copy commands without adapting them.
The commands are clearly present in Markdown code blocks for runbook examples. There is no script wrapper or automatic execution path in the skill file.
Static false positives ignored (2)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Low
Example URLs and IP Ranges Are Not Exfiltration
Verdict: FALSE_POSITIVE for network exfiltration. The hardcoded URLs point to example dashboards, status pages, documentation, local Prometheus, and placeholder company APIs. The IP ranges appear in an example Kubernetes NetworkPolicy, not in active network code.
The URLs and IP ranges are embedded in documentation examples and resource links. No code sends secrets or collected data to these endpoints.
Low
Weak Cryptography Alerts Are Substring False Positives
Verdict: FALSE_POSITIVE. The weak-cryptography static findings map to ordinary words or SQL ordering text in Markdown content, not cryptographic algorithm use. No evidence found of MD5, DES, SHA1, or custom cryptographic implementation.
The cited lines contain frontmatter text or SQL ORDER BY clauses. They do not define or invoke cryptographic primitives.

Detected Patterns

Copy-Paste Production Change Commands
Audited by: codex

Jul 1, 2026, 01:14 AM

Static analysis flagged many command, URL, IP address, filesystem, and weak-crypto patterns in SKILL.md. Review found these are Markdown runbook examples and false positives, not executable skill code or malicious behavior. The skill is safe to publish with a low-risk warning because operators may copy impactful production commands.

1
Files scanned
384
Lines analyzed
4
Review items
2
False positives ignored
Capability review items (1)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Low
Operational Command Examples Require Review
Verdict: TRUE_POSITIVE as content risk, not executable malware. SKILL.md contains Kubernetes, curl, psql, and shell command examples for incident response. They are inside Markdown templates, so the skill does not execute them, but users could cause production changes if they copy commands without adapting them.
The commands are clearly present in Markdown code blocks for runbook examples. There is no script wrapper or automatic execution path in the skill file.
Static false positives ignored (2)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Low
Example URLs and IP Ranges Are Not Exfiltration
Verdict: FALSE_POSITIVE for network exfiltration. The hardcoded URLs point to example dashboards, status pages, documentation, local Prometheus, and placeholder company APIs. The IP ranges appear in an example Kubernetes NetworkPolicy, not in active network code.
The URLs and IP ranges are embedded in documentation examples and resource links. No code sends secrets or collected data to these endpoints.
Low
Weak Cryptography Alerts Are Substring False Positives
Verdict: FALSE_POSITIVE. The weak-cryptography static findings map to ordinary words or SQL ordering text in Markdown content, not cryptographic algorithm use. No evidence found of MD5, DES, SHA1, or custom cryptographic implementation.
The cited lines contain frontmatter text or SQL ORDER BY clauses. They do not define or invoke cryptographic primitives.

Detected Patterns

Copy-Paste Production Change Commands
Audited by: codex

Jan 17, 2026, 09:28 AM

Pure documentation skill containing only static markdown templates. No executable code, no file system access, no network calls. All 70 static findings are false positives: markdown code blocks were misidentified as Ruby/shell backticks, example URLs were flagged as hardcoded network addresses, and SQL/YAML content was misidentified as weak cryptographic algorithms. The skill generates text templates only.

2
Files scanned
589
Lines analyzed
3
Review items
0
False positives ignored
Audited by: claude

Jan 17, 2026, 09:28 AM

Pure documentation skill containing only static markdown templates. No executable code, no file system access, no network calls. All 70 static findings are false positives: markdown code blocks were misidentified as Ruby/shell backticks, example URLs were flagged as hardcoded network addresses, and SQL/YAML content was misidentified as weak cryptographic algorithms. The skill generates text templates only.

2
Files scanned
589
Lines analyzed
3
Review items
0
False positives ignored
Audited by: claude

Jan 4, 2026, 04:31 PM

Pure documentation skill with static markdown templates. No executable code, no file access, no network calls. Contains example commands for user reference only. Low risk due to embedded example endpoints that require manual copy-paste to execute.

4
Files scanned
633
Lines analyzed
2
Review items
0
False positives ignored
Capability review items (1)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Low
Example network requests in templates
The templates include example network calls such as `curl -s "http://prometheus:9090/api/v1/query?query=sum(rate(http_requests_total{status=~'5..'}[5m]))"`. These are not executed by the skill, but could be copied and run against internal services. Users must review and customize endpoints before use.

Risk Factors

🌐 Network access (2)
Audited by: claude

Jan 4, 2026, 04:31 PM

Pure documentation skill with static markdown templates. No executable code, no file access, no network calls. Contains example commands for user reference only. Low risk due to embedded example endpoints that require manual copy-paste to execute.

4
Files scanned
633
Lines analyzed
2
Review items
0
False positives ignored
Capability review items (1)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Low
Example network requests in templates
The templates include example network calls such as `curl -s "http://prometheus:9090/api/v1/query?query=sum(rate(http_requests_total{status=~'5..'}[5m]))"`. These are not executed by the skill, but could be copied and run against internal services. Users must review and customize endpoints before use.

Risk Factors

🌐 Network access (2)
Audited by: claude