Security audit

Work Productivity Self Improving Workflow Helper

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only workflow helper with overly broad activation wording, but no hidden code, persistence, credential access, or automatic actions.

This skill is reasonable to install if you want a planning helper for agent reliability and self-improvement workflows. Be aware that implicit invocation is enabled with broad trigger words, so prefer explicit invocation or narrow the triggers if you manage the skill configuration.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (12)

Vague Triggers

High

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The description says to use the skill when a user asks for broad concepts like "self," "improving," or "logs," which is ambiguous invocation guidance rather than a precise scope boundary. In a skill-routing system, this can lead to accidental invocation across many benign prompts, creating prompt-surface expansion and making the system easier to manipulate through ordinary language rather than explicit capability requests.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The trigger list includes very common terms such as "self," "improving," "logs," and "own," which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that this skill hijacks routing or injects irrelevant workflow guidance into contexts where another skill or the base assistant should respond, reducing predictability and potentially interfering with safer task handling.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger keywords include very broad everyday terms such as "self", "own", and "improved", which can cause the skill to activate in many unrelated conversations. Overbroad activation can route user requests into the wrong workflow, causing prompt/context injection into unrelated tasks, confused agent behavior, or accidental execution of unsuitable guidance.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: Enabling implicit invocation without strict trigger constraints allows the skill to activate from loosely related natural language, increasing the chance of untrusted or accidental routing into this skill. In a self-improving workflow context, that is more dangerous because the skill may influence debugging, safety, setup, or workflow changes without clear user consent, potentially amplifying bad guidance across agent behavior.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger examples are phrased so broadly that ordinary user requests like asking for a practical workflow or bug-fixing help could activate this skill even when the user did not intend to invoke a self-improving agent workflow. Overbroad activation can misroute requests, expand the skill’s operational scope beyond its intended domain, and increase the chance that a more privileged or automation-oriented workflow is applied inappropriately.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The trigger sentences are generic and match broad everyday phrases like 'help me' and 'I need a practical workflow,' which can cause the skill to activate outside its intended scope. In an agent environment, overly broad activation boundaries can lead to unintended invocation, prompt/context hijacking opportunities, and incorrect routing of sensitive or unrelated tasks into this skill.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger keywords and example invocations are broad enough to match many ordinary requests, including generic terms like 'self', 'improving', 'logs', 'own', and 'bug fix'. This can cause the skill to activate unexpectedly, increasing prompt-surface exposure and creating routing confusion where unrelated user requests are handled by this skill without clear user intent.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The example trigger sentences model activation from very general wording and repeat broad requirement text instead of showing clear, constrained invocation examples. This reinforces loose matching behavior and increases the likelihood that downstream maintainers or automated systems will treat vague language as sufficient to invoke the skill.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The activation condition in the description is ambiguous because it mixes broad topical words with general requests for "practical workflow, artifact, checklist, analysis, or implementation support" without clear boundaries. This makes the skill easier to invoke outside its intended scope, increasing misrouting risk and reducing reliability of downstream agent behavior.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The default prompt includes very broad trigger terms such as 'work-productivity', 'self-improving-agent', and generic help wording, which are likely to overlap with ordinary user requests. This creates a prompt-squatting risk where the skill may be invoked unintentionally, causing unrelated conversations to be steered by the skill and exposing users to unexpected behavior or instruction injection through the skill content.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The keyword list includes extremely generic terms such as 'self', 'improving', 'logs', and 'improved', which are common across benign everyday conversations and unrelated tasks. This creates a high risk of accidental skill invocation, causing prompt/scope injection at the routing layer where the wrong skill may take over a session and produce actions or guidance outside the user's actual intent.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: This Chinese README contains substantial English instructional content and English trigger phrases, but it does not tell users they can choose their preferred language or that English is required. That can create a language-policy issue if users expect the zh-CN file to operate fully in Chinese by default.
Content: No source excerpt is available for this finding.