All 48 static alerts are false positives caused by Markdown formatting, fixed command substitution, or intentional Git worktree paths. Semantic review found real operational risks from destructive resets, remote history rewriting, and test execution across historical revisions. No prompt injection or malicious intent was found.
Executable hard reset examples can irreversibly discard uncommitted files and index changes without requiring a clean status check or explicit confirmation.
Each cited line directly invokes git reset --hard. The surrounding guidance does not consistently require preflight checks or confirmation.
Automated bisect runs repository scripts across historical revisions, which may execute untrusted code with the current user's credentials and network access.
Both cited commands instruct Git to execute project test commands repeatedly after checking out historical commits. No isolation or script-review warning is provided.
All static findings were adjudicated as false positives. They come from Markdown inline code, fenced Bash examples, and Git worktree paths rather than executable scripts, hidden automation, or prompt-injection content. The skill teaches destructive Git operations, so users should require repository-specific confirmation before changing history.
The static findings are false positives from Markdown inline code and fenced Git command examples. No prompt injection, data exfiltration, hidden network behavior, or executable helper code was found in SKILL.md. The examples include powerful Git operations, but they are transparent domain guidance.
Static analysis flagged many command and path patterns, but review found they are Markdown examples for legitimate Git workflows, not executable bundled code. No prompt injection, exfiltration, network access, or confirmed malicious intent was found. The skill still deserves a medium warning because several examples can rewrite history, discard local changes, or execute repository scripts if used without user confirmation.
1
Files scanned
413
Lines analyzed
4
Review items
3
False positives ignored
Capability review items (2)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The skill documents rebase, force push, and hard reset workflows. These are legitimate Git operations, but they can rewrite shared history or discard local work if executed without review.
The commands are visible in Markdown examples and directly include history rewrite and hard reset operations. They are not malicious by themselves, but the operational risk is real.
Repository Script Execution Through Bisect Examples
The skill documents automated bisect commands that run local scripts or test commands. This is valid Git usage, but it may execute untrusted repository code in the user environment.
The documented commands invoke ./test.sh and npm test through git bisect run. This is expected behavior for bisect automation, but it carries execution risk when the repository is untrusted.
Static false positives ignored (3)
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.
The Ruby and shell backtick detections are Markdown formatting and fenced bash examples. No Ruby file, executable script body, or hidden command runner was found in the scanned file.
The flagged text appears inside Markdown inline code or bash code blocks. There is no evidence that the skill executes these strings automatically.
The ../ paths are examples for git worktree directories beside the current project. They do not read arbitrary files or construct paths from user-controlled input.
The path patterns are fixed example directories used by git worktree commands. No filesystem traversal code or file exfiltration behavior is present.
The weak cryptography alerts do not correspond to cryptographic code. The flagged lines contain descriptive prose and best practice text for commit messages.
No hash function, encryption call, credential handling, or cryptographic configuration appears on the flagged lines. The alerts are scanner noise.
All static findings are false positives. The skill is a documentation guide for Git commands - the flagged 'external_commands' are Git examples in markdown code blocks, 'filesystem' findings are legitimate worktree relative paths, and 'weak cryptographic algorithm' was a misidentification. This is safe, legitimate educational content.
1
Files scanned
415
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.