$ loading_
Detects rumination, hyperfocus, and sycophancy risks without silently blocking interactions.
The available material is very limited, but based on the facts provided, this MCP tool requires no secrets, declares no remote endpoints, and comes from an official registry as an open-source project, so the overall risk appears low. The main point to watch is local code execution inherent to an MCP tool, while the missing README limits verification of its exact behavior and data access scope.
The material explicitly states that no keys or environment variables are required, and no API tokens, account credentials, or other sensitive secrets are mentioned, so credential exposure or misuse risk appears low.
No remote endpoints or external hosts are declared, and the provided material does not indicate that user data is sent to third-party services; based on known facts, there is no clear data egress path.
The objective checks indicate an executes-code capability, meaning it runs code/processes locally as an MCP tool. This is normal for this class of tool, but it still warrants least-privilege execution and isolation of its runtime environment.
The documentation contains only a one-line description and no README, so it does not clearly state what local files, prompts, or session data it may read or write. There is no evidence of overbroad access, but the data access boundary is insufficiently documented and should be verified in practice.
Positive factors include distribution via an official registry, an auditable open-source repository, and updates within the last year; these materially reduce risk. Points to watch are the unspecified license, very low community adoption, and sparse documentation, which limit overall supply-chain confidence.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "NeuroDock Guardrail" MCP server from askskill: Run: claude mcp add 'io-github-tlennon-ie-neurodock-mcp-guardrail' -- uvx neurodock-mcp-guardrail
Check whether the following support bot reply shows rumination, hyperfocus, or sycophancy. Provide severity-based warnings, but do not block the output: {{reply_text}}Returns risk types, trigger reasons, severity levels, and improvement suggestions while preserving the original output.
Act as a conversational guardrail and analyze this model draft for behavioral risks: repeated rumination, excessive fixation on one goal, or blindly agreeing with the user. Output advisory warnings and do not silently block it. Text: {{draft_response}}Generates structured advisory results that can be shown in the product UI or logged for audits.
Evaluate this test prompt and model answer to determine whether they trigger sycophancy, rumination, or hyperfocus. Explain the signals and suggest fixes: Prompt: {{prompt}}; Response: {{response}}Outputs a testing report identifying problematic patterns and actionable prompt-improvement recommendations.
Local-first cognitive support with memory, time awareness, and communication guardrails.
Break vague goals into actionable tasks and reveal the next safe step.
Decode message subtext, rewrite replies, and summarize meetings clearly.
Build a local cognitive graph linking people, projects, and decisions.
Provides time cues, session boundaries, and break prompts for time-blind users.
Enforce compliant AI agent tool calls with deterministic decisions and audit trails.
Run anti-collusion detection to identify possible collusion risk signals.
Prevent destructive actions in production systems and autonomous agent workflows.
Score agent outputs with guardrails, policy checks, injection, and PII detection.
Detects agent infinite loops and provides safeguards with recovery recommendations.
Protect MCP-connected agents with PII redaction, rate limits, and policy enforcement.