Skill

performing-serverless-function-security-review

skill:mukul975/anthropic-cybersecurity-skills#performing-serverless-function-security-review

Performing security reviews of serverless functions across AWS Lambda, Azure Functions, and GCP Cloud Functions to identify overly permissive execution roles, insecure environment variables, injection vulnerabilities, and missing runtime protections.

SkillsMP

Verdict

Suspicious

Risky behavior the analysis could not rule out.

Analyzed Oct 27, 2026, 23:23 UTC at 54a7988

findings

medium

highest severity

threat classes

4,532

files scanned

96/96

checks satisfied

package advisories

Why this verdict

suspicious: finding skill:skills/performing-serverless-function-security-review:skills/performing-serverless-function-security-review/SKILL.md:204:SUPPLY

Security findings 1

Supply chain

mediumContextual riskmedium confidence

The skill's Step 5 code-review workflow instructs the agent to install unpinned, mutable remote code from public registries at run time (pip install bandit; npm install -g eslint @microsoft/eslint-plugin-sdl), which is an install-time remote-code fetch the declared description does not mention, though the packages are official tools serving the declared analysis purpose.

skills/performing-serverless-function-security-review/SKILL.md:204-208

pip install bandit

vector:instruction

Counter-argument considered

Strongest refutation case: the packages (bandit, eslint, @microsoft/eslint-plugin-sdl) are official security scanners serving the declared purpose of identifying injection vulnerabilities, suggesting a declared-core-capability carve-out; however, the carve-out for installs requires pinning and these bash commands give no version pins, hashes, or lockfile, so the run-time mutable fetch from public registries stands undeclared by the description (which names no tool installation) and is concretely instructed to the agent, so the finding survives.