📦
Audit History
secrets-management - 8 audits
Version comparison
Capability and finding changes across audited versions, newest first.
| Version | Date | Result | Review items | Change vs previous |
|---|---|---|---|---|
| v8 Latest | Jul 8, 2026, 01:50 PM | 3 confirmed | 6 | No capability change |
| v7 | Jul 8, 2026, 01:50 PM | 3 confirmed | 6 | No capability change |
| v6 | Jul 1, 2026, 01:01 AM | No confirmed findings | 3 | No capability change |
| v5 | Jul 1, 2026, 01:01 AM | No confirmed findings | 3 | No capability change |
| v4 | Jan 17, 2026, 10:01 AM | No confirmed findings | 0 | No capability change |
| v3 | Jan 17, 2026, 10:01 AM | No confirmed findings | 0 | Network accessExternal commandsEnv variables |
| v2 | Jan 4, 2026, 04:22 PM | No confirmed findings | 0 | No capability change |
| v1 | Jan 4, 2026, 04:22 PM | No confirmed findings | 0 | Baseline |
Audit version 8 Latest
Jul 8, 2026, 01:50 PM
The skill is documentation-only and contains no prompt injection attempt or runtime code. Most command, URL, and environment detections are false positives from fenced examples, but several examples teach unsafe secret handling through CI log output, Terraform state persistence, and weak demo credentials.
1
Files scanned
347
Lines analyzed
12
Review items
0
False positives ignored
Confirmed security concerns (3)
High
Secret Values Printed to CI Logs
Several CI examples use echo with API keys or database URLs. This teaches users to place secret values in build logs, where they may be retained or shared.
The examples explicitly print secret variables using echo. This is direct evidence of a log disclosure pattern.
High
Terraform State May Store Secret Values
The Terraform example reads a database password from AWS Secrets Manager and assigns it to an RDS password field. Terraform state can retain that value.
The snippet shows the secret value assigned into Terraform-managed infrastructure. Terraform state commonly records resource arguments unless a provider handles them as write-only.
Medium
Insecure Demo Credentials Lack Strong Warning
The Vault setup example uses a root token and a literal password value. It is labeled as a dev server, but the snippet needs a clearer non-production warning.
The context says dev server, which reduces severity. The root token and simple password are still risky if copied into reusable automation.
Capability review items (6)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
High
Database connection strings
password = jsondecode(data.aws_secretsmanager_secret_version.db_password.secret_string)["
The Terraform example passes a retrieved database password into a resource argument, which can persist secret material in Terraform state. This is a real exposure risk even though the value comes from Secrets Manager.
High
Database connection strings
echo "Database URL: ${{ secrets.DATABASE_URL }}"
The CI example echoes a database URL secret into job logs. Masking can fail for derived or unconfigured values, so this encourages credential disclosure.
High
Database connection strings
- echo "Database: $DATABASE_URL"
The GitLab example echoes a database URL variable into job logs. This can disclose credentials embedded in the connection string.
High
Generic API/secret keys
echo "API Key: ${{ secrets.API_KEY }}"
The GitHub Actions example prints an API key secret with echo. Even with masking, examples should never instruct users to write secret values to logs.
High
Generic API/secret keys
echo "Deploying with ${{ secrets.PROD_API_KEY }}"
The environment secret example prints a production API key in a deployment log line. This is a direct secret disclosure pattern.
High
Generic API/secret keys
- echo "Deploying with $API_KEY"
The GitLab variable example echoes an API key into job output. CI logs are commonly retained and shared, so this is a real leakage risk.
Risk Factors
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
🌐 Network access (5)
Audited by: codex
Jul 8, 2026, 01:50 PM
The skill is documentation-only and contains no prompt injection attempt or runtime code. Most command, URL, and environment detections are false positives from fenced examples, but several examples teach unsafe secret handling through CI log output, Terraform state persistence, and weak demo credentials.
1
Files scanned
347
Lines analyzed
12
Review items
0
False positives ignored
Confirmed security concerns (3)
High
Secret Values Printed to CI Logs
Several CI examples use echo with API keys or database URLs. This teaches users to place secret values in build logs, where they may be retained or shared.
The examples explicitly print secret variables using echo. This is direct evidence of a log disclosure pattern.
High
Terraform State May Store Secret Values
The Terraform example reads a database password from AWS Secrets Manager and assigns it to an RDS password field. Terraform state can retain that value.
The snippet shows the secret value assigned into Terraform-managed infrastructure. Terraform state commonly records resource arguments unless a provider handles them as write-only.
Medium
Insecure Demo Credentials Lack Strong Warning
The Vault setup example uses a root token and a literal password value. It is labeled as a dev server, but the snippet needs a clearer non-production warning.
The context says dev server, which reduces severity. The root token and simple password are still risky if copied into reusable automation.
Capability review items (6)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
High
Database connection strings
password = jsondecode(data.aws_secretsmanager_secret_version.db_password.secret_string)["
The Terraform example passes a retrieved database password into a resource argument, which can persist secret material in Terraform state. This is a real exposure risk even though the value comes from Secrets Manager.
High
Database connection strings
echo "Database URL: ${{ secrets.DATABASE_URL }}"
The CI example echoes a database URL secret into job logs. Masking can fail for derived or unconfigured values, so this encourages credential disclosure.
High
Database connection strings
- echo "Database: $DATABASE_URL"
The GitLab example echoes a database URL variable into job logs. This can disclose credentials embedded in the connection string.
High
Generic API/secret keys
echo "API Key: ${{ secrets.API_KEY }}"
The GitHub Actions example prints an API key secret with echo. Even with masking, examples should never instruct users to write secret values to logs.
High
Generic API/secret keys
echo "Deploying with ${{ secrets.PROD_API_KEY }}"
The environment secret example prints a production API key in a deployment log line. This is a direct secret disclosure pattern.
High
Generic API/secret keys
- echo "Deploying with $API_KEY"
The GitLab variable example echoes an API key into job output. CI logs are commonly retained and shared, so this is a real leakage risk.
Risk Factors
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
🌐 Network access (5)
Audited by: codex
Jul 1, 2026, 01:01 AM
Static analysis correctly found many command, network, and credential patterns, but they appear inside SKILL.md documentation examples rather than executable skill code. No prompt injection or malicious exfiltration intent was found, but several examples encourage printing secrets into CI logs and should carry a marketplace warning or be fixed.
1
Files scanned
347
Lines analyzed
6
Review items
2
False positives ignored
Capability review items (3)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Medium
Secret Values Printed in CI Logs
The documentation includes CI examples that echo API keys, database URLs, and production API keys to job logs. This is dangerous because users may copy the examples and expose secrets through CI output even when platform secret stores are used.
The cited lines explicitly print variables named API key, database URL, and production API key in CI run steps. The context confirms this is documentation, not malicious code, but the unsafe recommendation is clear.
Medium
Hardcoded Demo Secrets in Examples
The documentation uses a Vault root token, a literal password value, and a literal AWS secret string in examples. These are presented as setup snippets, but copying them into real environments would create weak or shared credentials.
The examples contain concrete secret-like values and a root token. The Vault block is labeled as a dev server, which lowers maliciousness but does not remove the copy-paste risk.
Low
Network Indicators Use Placeholder or Local Endpoints
The hardcoded URLs and IP address point to localhost or example Vault endpoints in setup documentation. No evidence found of a hidden external collection endpoint or data exfiltration workflow.
The URLs are standard Vault examples and include localhost or example.com hostnames. They do not appear to be attacker-controlled endpoints.
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
Command Execution Findings Are Documentation Examples
The shell, Vault, AWS CLI, Docker, Terraform, and CI command patterns are inside Markdown code fences. No evidence found that the skill contains executable scripts or automatically runs these commands.
The line context shows fenced documentation blocks rather than runnable repository files. The commands are still relevant risk indicators because users may execute them manually.
Low
Weak Cryptography Static Finding Appears False Positive
The weak cryptographic algorithm alerts correspond to text in the skill description and related skill name, not to cryptographic code. No evidence found of MD5, DES, RC4, or similar algorithms being recommended or executed.
The cited lines are metadata text and a related-skill bullet. They do not contain crypto implementation guidance.
Risk Factors
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
🌐 Network access (5)
Detected Patterns
CI Examples Echo Secret VariablesCopy-Paste Shell Commands Handle Live Secrets
Audited by: codex
Jul 1, 2026, 01:01 AM
Static analysis correctly found many command, network, and credential patterns, but they appear inside SKILL.md documentation examples rather than executable skill code. No prompt injection or malicious exfiltration intent was found, but several examples encourage printing secrets into CI logs and should carry a marketplace warning or be fixed.
1
Files scanned
347
Lines analyzed
6
Review items
2
False positives ignored
Capability review items (3)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Medium
Secret Values Printed in CI Logs
The documentation includes CI examples that echo API keys, database URLs, and production API keys to job logs. This is dangerous because users may copy the examples and expose secrets through CI output even when platform secret stores are used.
The cited lines explicitly print variables named API key, database URL, and production API key in CI run steps. The context confirms this is documentation, not malicious code, but the unsafe recommendation is clear.
Medium
Hardcoded Demo Secrets in Examples
The documentation uses a Vault root token, a literal password value, and a literal AWS secret string in examples. These are presented as setup snippets, but copying them into real environments would create weak or shared credentials.
The examples contain concrete secret-like values and a root token. The Vault block is labeled as a dev server, which lowers maliciousness but does not remove the copy-paste risk.
Low
Network Indicators Use Placeholder or Local Endpoints
The hardcoded URLs and IP address point to localhost or example Vault endpoints in setup documentation. No evidence found of a hidden external collection endpoint or data exfiltration workflow.
The URLs are standard Vault examples and include localhost or example.com hostnames. They do not appear to be attacker-controlled endpoints.
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
Command Execution Findings Are Documentation Examples
The shell, Vault, AWS CLI, Docker, Terraform, and CI command patterns are inside Markdown code fences. No evidence found that the skill contains executable scripts or automatically runs these commands.
The line context shows fenced documentation blocks rather than runnable repository files. The commands are still relevant risk indicators because users may execute them manually.
Low
Weak Cryptography Static Finding Appears False Positive
The weak cryptographic algorithm alerts correspond to text in the skill description and related skill name, not to cryptographic code. No evidence found of MD5, DES, RC4, or similar algorithms being recommended or executed.
The cited lines are metadata text and a related-skill bullet. They do not contain crypto implementation guidance.
Risk Factors
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
🌐 Network access (5)
Detected Patterns
CI Examples Echo Secret VariablesCopy-Paste Shell Commands Handle Live Secrets
Audited by: codex
Jan 17, 2026, 10:01 AM
Documentation-only skill providing CI/CD secrets management guidance. Static findings triggered on legitimate tool references (Vault, AWS), environment variable examples, and shell script documentation - all expected content for a secrets management educational resource. No executable code or malicious patterns detected.
2
Files scanned
524
Lines analyzed
3
Review items
0
False positives ignored
Risk Factors
🌐 Network access (5)
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
Audited by: claude
Jan 17, 2026, 10:01 AM
Documentation-only skill providing CI/CD secrets management guidance. Static findings triggered on legitimate tool references (Vault, AWS), environment variable examples, and shell script documentation - all expected content for a secrets management educational resource. No executable code or malicious patterns detected.
2
Files scanned
524
Lines analyzed
3
Review items
0
False positives ignored
Risk Factors
🌐 Network access (5)
⚙️ External commands (40)
SKILL.md:52-65 SKILL.md:65-69 SKILL.md:69-94 SKILL.md:94-98 SKILL.md:98-111 SKILL.md:111-113 SKILL.md:113-119 SKILL.md:119-123 SKILL.md:123-127 SKILL.md:127-148 SKILL.md:148-152 SKILL.md:152-164 SKILL.md:164-170 SKILL.md:170-175 SKILL.md:175-179 SKILL.md:179-187 SKILL.md:187-189 SKILL.md:189-195 SKILL.md:195-200 SKILL.md:200-224 SKILL.md:224-251 SKILL.md:251-265 SKILL.md:265-305 SKILL.md:305-311 SKILL.md:311-324 SKILL.md:324-328 SKILL.md:328-335 SKILL.md:335-339 SKILL.md:339-340 SKILL.md:340-344 SKILL.md:344-345 SKILL.md:345-346 SKILL.md:107 SKILL.md:108 SKILL.md:137-140 SKILL.md:316 SKILL.md:98-111 SKILL.md:127-148 SKILL.md:311-324 SKILL.md:312
Audited by: claude
Jan 4, 2026, 04:22 PM
Only documentation content was found, with examples for secrets tools and CI usage. No code paths show data theft, malicious execution, or hidden network activity.
4
Files scanned
353
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Audited by: claude
Jan 4, 2026, 04:22 PM
Only documentation content was found, with examples for secrets tools and CI usage. No code paths show data theft, malicious execution, or hidden network activity.
4
Files scanned
353
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Audited by: claude