Back to skill

Security audit

Pptx Master V1.2.3 20260507

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a coherent PPT-generation toolkit, but it includes high-impact update and networking behaviors that need review before installation.

Review this skill before installing. Use it only in a contained project environment, avoid running scripts/update_repo.py unless you inspect the pulled changes first, prefer --skip-pip for updates, keep API keys scoped and low-quota, and do not use custom image-provider base URLs unless you trust the endpoint. Be cautious converting sensitive HTTPS pages because one converter disables certificate verification.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (4)

T03 · Remote Payload Retrieval and Execution

Error
Location
scripts/update_repo.py:84
Finding

Mutable Repository Update Followed by Automatic Dependency Installation

Content
View full analysis
None: if not REQUIREMENTS_FILE.exists(): print("requirements.txt not found; skipping Python dependency sync.") return print("requirements.txt changed. Syncing Python dependencies...") result = run_command([sys.executable, "-m", "pip", "install", "-r", str(REQUIREMENTS_FILE)]) if result.stdout.strip(): print(result.stdout.strip()) if result.stderr.strip(): print(result.stderr.strip()) def main() -> int: args = parse_args() try: ensure_git_available() ensure_clean_tracked_worktree() before_head = get_head_revision() before_requirements = file_digest(REQUIREMENTS_FILE) print(f"Repository: {REPO_ROOT}") pull_result = run_command(["git", "pull", "--ff-only"]) if pull_result.stdout.strip(): print(pull_result.stdout.strip()) if pull_result.stderr.strip(): print(pull_result.stderr.strip()) after_head = get_head_revision() after_requirements = file_digest(REQUIREMENTS_FILE) if before_head == after_head: print("Repository is already up to date.") else: print(f"Updated from {before_head[:7]} to {after_head[:7]}.") if args.skip_pip: print("Skipped Python dependency sync (--skip-pip).") elif before_requirements != after_requirements: sync_python_dependencies() else: print("requirements.txt unchanged. Skipping Python dependency sync.") ``` The operation is explicitly exposed as a normal repository-maintenance command: ```markdown Repository update: ```bash python3 scripts/update_ ...[truncated 2318 chars]
Remediation
View remediation

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:42
Finding

Skill Instruction Precedence Hijacking

Content
View full analysis
## 🚨 Global Execution Discipline (MANDATORY) > > **This workflow is a strict serial pipeline. The following rules have the highest priority — violating any one of them constitutes execution failure:** > > 1. **SERIAL EXECUTION** — Steps MUST be executed in order; the output of each step is the input for the next. Non-BLOCKING adjacent steps may proceed continuously once prerequisites are met, without waiting for the user to say "continue" > 2. **BLOCKING = HARD STOP** — Steps marked ⛔ BLOCKING require a full stop; the AI MUST wait for an explicit user response before proceeding and MUST NOT make any decisions on behalf of the user > 3. **NO CROSS-PHASE BUNDLING** — Cross-phase bundling is FORBIDDEN. (Note: the Eight Confirmations in Step 4 are ⛔ BLOCKING — the AI MUST present recommendations and wait for explicit user confirmation before proceeding. Once the user confirms, all subsequent non-BLOCKING steps — design spec output, SVG generation, speaker notes, and post-processing — may proceed automatically without further user confirmation) > 4. **GATE BEFORE ENTRY** — Each Step has prerequisites (🚧 GATE) listed at the top; these MUST be verified before starting that Step > 5. **NO SPECULATIVE EXECUTION** — "Pre-preparing" content for subsequent Steps is FORBIDDEN (e.g., writing SVG code during the Strategist phase) ``` A separate instruction extends this precedence claim to conflicts with other skills: ```markdown - If another generic coding skill suggests repository conventions that conflict with this workflow, follow this skill first unless the user explicitly asks otherwise ``` ### Technical Analysis A Skill may define its own operational workflow, but it must not claim global or highest instruction priority. The phrases “highest priority” ...[truncated 1330 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/web_to_md.py:80
Finding

HTTPS Certificate Verification Disabled for Web and Image Downloads

Content
View full analysis
str: """Fetch a web page with explicit headers and encoding detection. Args: url: Target URL. Returns: The response body as text. """ headers = { "User-Agent": CONFIG["user_agent"], "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8", "Accept-Language": "zh-CN,zh;q=0.9,en;q=0.8" } try: response = requests.get(url, headers=headers, timeout=CONFIG["timeout"], verify=False) response.raise_for_status() # Enhanced encoding detection (requests handles this well usually, but we force apparent_encoding for Chinese) response.encoding = response.apparent_encoding return response.text except Exception as e: raise Exception(f"Failed to fetch {url}: {str(e)}") ``` The same insecure setting is used when retrieving embedded images: ```python else: try: resp = requests.get( abs_url, headers={"User-Agent": CONFIG["user_agent"]}, timeout=CONFIG["timeout"], verify=False, ) resp.raise_for_status() filename = build_image_filename( abs_url, idx, resp.headers.get("Content-Type")) ``` ### Technical Analysis Passing `verify=False` disables validation of the server certificate and hostname for HTTPS requests. Encryption may still occur, but the client no longer establishes that it is communicating with the intended server. This affects both primary webpage retrieval and secondary image downloads. As a result, an attacker able to intercept traffic can present an arbitrary cer ...[truncated 1451 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/image_backends/backend_stability.py:62
Finding

Provider Credentials and Confidential Prompts Can Be Forwarded to Unrestricted Custom Endpoints

Content
View full analysis
str: """Generate one image with the Stability backend.""" if aspect_ratio not in VALID_ASPECT_RATIOS: raise ValueError( f"Unsupported aspect ratio '{aspect_ratio}' for Stability backend. " f"Supported: {VALID_ASPECT_RATIOS}" ) resolved_model, url = _resolve_endpoint(model, image_size, base_url) headers = { "Authorization": f"Bearer {api_key}", "Accept": "image/*", } data = { "prompt": prompt, "aspect_ratio": aspect_ratio, "output_format": "png", } if negative_prompt: data["negative_prompt"] = negative_prompt print("[Stability AI]") print(f" Model: {resolved_model}") print(f" Prompt: {prompt[:120]}{'...' if len(prompt) > 120 else ''}") print(f" Aspect Ratio: {aspect_ratio}") print(f" Preset Size: {image_size}") print() p ...[truncated 2856 chars]
Remediation
View remediation
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (632)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill description and workflow claim a broad PPTX-generation pipeline, but the findings indicate substantial undeclared behaviors in underlying code, including external API use, local file rewriting, image download/save, and unrelated utilities. This mismatch is dangerous because agents and users may grant trust and permissions based on the declared purpose while hidden or unrelated capabilities execute under that trust boundary.

Content

No source excerpt is available for this finding.