tokportal MCP: Real-Time TikTok and Instagram Shadowban Diagnostics in Claude and Astra
tokportal has rolled out an MCP update enabling Claude and Astra users to connect TikTok and Instagram accounts, diagnose shadowbans with real account metrics,
Social media automation platform tokportal has released a major update to the 'tokportal MCP', enabling AI clients like Anthropic Claude and Astra to connect directly to TikTok and Instagram accounts via the Model Context Protocol. Rather than forcing creators and engineering teams to rely on guesswork or anecdotal evidence to determine whether an account has suffered a shadowban, the updated tool feeds verified account-level diagnostic metrics into the AI context window to evaluate real reach, content flags, and algorithmic distribution health.
Image source: @tellenne_ / Vincent (tokportal) via @nicos_ai (X)
When impressions drop abruptly on short-form video platforms, determining whether a channel has been penalized by moderation algorithms or simply failed to engage viewers has historically been difficult. By bridging social media analytics into modern desktop AI agent environments, the tokportal MCP provides developers and content teams with the raw diagnostics required to audit moderation signals and orchestrate recovery strategies through standard conversational prompts.
Beyond Guesswork: Direct Metrics for Account Health and Shadowbans
The primary strength of the tokportal MCP lies in replacing subjective assumptions with direct account-level data points exposed to LLMs. While traditional third-party analytics tools often guess penalties based strictly on sudden view drops, tokportal provides access to concrete operational indicators directly from the account.
- Real-Time Restriction Checks: Directly inspect account-level flags, soft-locks, temporary distribution halts, and suspension statuses.
- Authentic Shadowban Identification: Determine whether a channel's distribution is actively suppressed by platform filters, removing ambiguity around account status.
- Composite Account Health Scores: Evaluate overall standing across multiple operational dimensions that influence algorithmic recommendation weights.
Access to verified metrics enables teams to identify technical penalties immediately, preventing hasty account abandonments and unfocused re-upload loops that can further damage channel standing.
Video Distribution Status, AI Content Flags, and Deep Comment Analysis
In addition to channel-wide audits, the MCP provides micro-level diagnostics for each published asset, allowing creators to inspect how recommendation algorithms evaluate individual posts.
- Feed Distribution Tracking: Check whether specific clips are actively delivered to recommendation surfaces, such as the TikTok For You Feed, or if algorithmic restrictions are restricting reach.
- AI Content Detection Indicators: Identify whether platform automated moderation systems have tagged an asset as AI-generated media, alongside any associated labeling or reach modifiers.
- Sound Distribution and Comment Audits: Analyze background audio licensing impacts and extract comprehensive interaction data to understand how community engagement influences distribution.
Isolating issues at the single-video level allows creators to pinpoint exact policy or audio triggers before modifying subsequent production workflows.
Automated Self-Healing Loops and Operational Requirements in Claude and Astra
Beyond static reporting, the tokportal MCP is structured to support closed-loop automated workflows in clients such as Claude Desktop and Astra. Users can prompt their assistant to execute structured diagnostics and automated warming routines directly within the chat interface.
- Account Connection and Fleet Health Scan: Connect target TikTok and Instagram accounts to the MCP server to inspect overarching distribution statuses and recent video performance.
- Cross-Post Anomaly Auditing: When a flag or distribution restriction is detected on a specific video, the AI cross-references metadata, sounds, and visual structures against normally performing posts to isolate anomalies.
- Account Warming and Progressive Scaling: Execute automated warming protocols to rebuild account trust, continuously monitoring feed re-entry signals before resuming scaled content publication.
Operational prerequisites include an active tokportal account integration and an MCP-compliant host such as Claude Desktop or Astra. Teams should note that because social media algorithms and moderation policies evolve continuously, the speed and efficacy of account recovery workflows will vary depending on prevailing platform guidelines.