Nm Sanctum Pr Review

Scope-focused PR review with requirements validation and backlog triage

This skill appears to be a well-documented PR-review playbook, but before installing: (1) confirm the runtime environment provides the CLI tools it assumes (gh or glab, git, jq, python and standard Unix tools). (2) Understand which credentials will be used — gh/gh auth or environment tokens (GH_TOKEN/GITLAB_TOKEN) are required to post comments; ensure any token granted has minimal scopes (repo:status/comments/PR write as needed). (3) Review what the declared config paths (night-market.* and memory-palace) map to in your agent/platform — verify they do not expose additional secrets you don't intend to share. (4) If you need stricter auditing, ask the publisher to explicitly declare required binaries and credential variables and to explain where captured knowledge is stored and who can access it. These clarifications would change this assessment to benign if provided.

Type: OpenClaw Skill Name: nm-sanctum-pr-review Version: 1.0.0 The skill bundle provides a comprehensive PR review framework that relies heavily on executing shell commands via the GitHub (gh) and GitLab (glab) CLIs. While the functionality is aligned with the stated purpose of code review and backlog triage, the inclusion of complex shell pipelines in SKILL.md, modules/pr-hygiene.md, and modules/version-validation.md creates a significant attack surface. Specifically, the scripts process untrusted input from PR metadata (e.g., branch names, commit messages, and file contents) using tools like awk, sed, and grep, which could lead to shell injection or unintended command execution if the AI agent does not strictly sanitize the environment. No evidence of intentional malice, such as data exfiltration or backdoors, was detected.