Seven static findings are confirmed because the skill documents npx, pip, brew, or apt-get commands that can execute code or modify the host. The remaining findings are false positives from Markdown formatting, standard cache paths, a relative documentation link, or non-reconnaissance prose. No prompt injection, credential exfiltration, or hidden malicious intent was found in SKILL.md.
1
Files scanned
244
Lines analyzed
9
Review items
0
False positives ignored
Capability review items (7)
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 bash example instructs running npx hyperframes tts, which executes external Node package code and writes media output. This is expected functionality but requires user confirmation.
Use `--lang` only to override auto-detection (stylized accents). Valid codes: `en-us`, `en-gb`, `es`
The requirements text includes brew install espeak-ng and apt-get install espeak-ng, which modify the host system. This setup path is legitimate but needs explicit approval.
Python 3.8+ with `kokoro-onnx` and `soundfile` (`pip install kokoro-onnx soundfile`). Model download
The requirements text includes pip install kokoro-onnx soundfile, which changes the Python environment and downloads packages. This is expected setup behavior but security-relevant.
Seven static findings are confirmed because the skill documents npx, pip, brew, or apt-get commands that can execute code or modify the host. The remaining findings are false positives from Markdown formatting, standard cache paths, a relative documentation link, or non-reconnaissance prose. No prompt injection, credential exfiltration, or hidden malicious intent was found in SKILL.md.
1
Files scanned
244
Lines analyzed
9
Review items
0
False positives ignored
Capability review items (7)
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 bash example instructs running npx hyperframes tts, which executes external Node package code and writes media output. This is expected functionality but requires user confirmation.
Use `--lang` only to override auto-detection (stylized accents). Valid codes: `en-us`, `en-gb`, `es`
The requirements text includes brew install espeak-ng and apt-get install espeak-ng, which modify the host system. This setup path is legitimate but needs explicit approval.
Python 3.8+ with `kokoro-onnx` and `soundfile` (`pip install kokoro-onnx soundfile`). Model download
The requirements text includes pip install kokoro-onnx soundfile, which changes the Python environment and downloads packages. This is expected setup behavior but security-relevant.
Most static findings are false positives from Markdown backticks, inline command names, cache-path documentation, and a relative documentation link. The real security consideration is that the skill asks users to run local CLI, pip, and package-manager commands that may install dependencies, download models, and write media outputs.
1
Files scanned
244
Lines analyzed
9
Review items
0
False positives ignored
Capability review items (7)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
This fenced command example instructs users or agents to run npx hyperframes. The command is core skill functionality, but it executes an external CLI and may download models or write outputs.
Use `--lang` only to override auto-detection (stylized accents). Valid codes: `en-us`, `en-gb`, `es`
This line recommends package-manager commands for espeak-ng, which can modify the user system. The command is expected for setup, but it is still an external command risk.
Python 3.8+ with `kokoro-onnx` and `soundfile` (`pip install kokoro-onnx soundfile`). Model download
This line recommends pip installing Python packages and notes model caching. The setup is legitimate, but it executes external package installation in the user environment.
Most static findings are false positives from Markdown backticks, inline command names, cache-path documentation, and a relative documentation link. The real security consideration is that the skill asks users to run local CLI, pip, and package-manager commands that may install dependencies, download models, and write media outputs.
1
Files scanned
244
Lines analyzed
9
Review items
0
False positives ignored
Capability review items (7)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
This fenced command example instructs users or agents to run npx hyperframes. The command is core skill functionality, but it executes an external CLI and may download models or write outputs.
Use `--lang` only to override auto-detection (stylized accents). Valid codes: `en-us`, `en-gb`, `es`
This line recommends package-manager commands for espeak-ng, which can modify the user system. The command is expected for setup, but it is still an external command risk.
Python 3.8+ with `kokoro-onnx` and `soundfile` (`pip install kokoro-onnx soundfile`). Model download
This line recommends pip installing Python packages and notes model caching. The setup is legitimate, but it executes external package installation in the user environment.
Static analysis flagged many command, filesystem, and blocker patterns, but most are false positives from Markdown examples and media terminology. A real medium risk remains because the skill instructs agents to run local CLI commands that download models and process user files.
Local CLI Commands Download Models and Process Files
The skill directs agents to run npx hyperframes commands for TTS, transcription, and background removal. This is legitimate for the skill, but it executes local tools, downloads model files, and reads or writes user media paths.
The command examples are explicit and repeated. They appear purposeful for media processing rather than malicious, so the risk is operational rather than a confirmed compromise.
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.
Cache and Repository Paths Are Benign Filesystem Use
The hidden home-directory findings refer to documented model cache paths under ~/.cache/hyperframes. The traversal alert is a repository-relative documentation link, not evidence of unsafe path handling.
The cited lines are explanatory Markdown, cache locations, and a relative link. No code constructs or user-controlled path joins are present in the reviewed file.
Weak Cryptography and Reconnaissance Alerts Are Scanner Noise
The blocker findings match media terms such as model sizes, Whisper model names, IDs, CRF values, and HTML data attributes. No cryptographic API, hashing routine, system probing, or secret collection is shown.
The surrounding context is media workflow documentation. I found no evidence of weak cryptographic algorithms or reconnaissance behavior in SKILL.md.
All 144 static analysis findings are false positives. The scanned file (SKILL.md) is a markdown documentation file, not executable code. Backtick patterns flagged as 'Ruby/shell execution' are shell commands inside markdown code blocks showing CLI usage examples. Filesystem and hidden file path alerts reference legitimate model cache directories (~/.cache/hyperframes/). Weak cryptographic algorithm and system reconnaissance alerts are text matching errors against model names, language codes, and documentation references. The skill poses no security risk as it only instructs users to run CLI commands with no data exfiltration, network callbacks, or hidden behavior.
1
Files scanned
244
Lines analyzed
2
Review items
5
False positives ignored
Static false positives ignored (5)
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.
False positive. SKILL.md is a markdown documentation file. Backtick patterns flagged by the static analyzer are shell commands inside markdown fenced code blocks. These are CLI usage examples showing users how to run npx hyperframes commands for legitimate media preprocessing tasks. No executable code is present. This is standard practice for CLI tool documentation.
False positive. References to ~/.cache/hyperframes/ are legitimate model cache directory paths used by ML pipelines. Models (Kokoro TTS, Whisper, u2net) are downloaded on first run and cached locally. This is standard behavior for local ML tools and poses no security risk as only model weights are stored.
False positive. The '..' path at line 90 is a relative markdown link reference to '../hyperframes/references/transcript-guide.md'. This is a documentation cross-reference, not a path traversal attack. The pattern is a standard relative URL in markdown files.
False positive. The static analyzer misidentified model names (u2net, small.en, medium), version numbers, and language codes as cryptographic algorithm references. These are model identifiers and configuration values, not cryptographic primitives. No actual weak crypto usage exists in this documentation file.
False positive. References at line 41 and 128 are documentation about system requirements (espeak-ng dependency) and output format compatibility (webm/mov/png). These are informational notes, not reconnaissance commands. No system enumeration is performed.