视频索引：抽帧 + ASR 语音转写 + VLM 视觉分段理解。输出时间线。
AI agents call video_index to retrieve information from Local Mmcp without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
video_index performs local video analysis operations: sampling frames, transcribing audio, and applying vision-language understanding to generate indexed metadata (a timeline). No data is created, modified, deleted, or executed as side effects—only read and analyzed. The tool produces an output representation of the video's content but does not alter the source material or trigger external operations.
From the tool's definition Tool description indicates: 抽帧 (frame extraction) + ASR (speech-to-text transcription) + VLM (visual segmentation understanding) producing a timeline output. These are all information extraction and analysis operations with no modifications to the video data.
Attacks that exploit this kind of access
视频索引：抽帧 + ASR 语音转写 + VLM 视觉分段理解。输出时间线。. It is categorised as a Read tool in the Local Mmcp MCP Server, which means it retrieves data without modifying state.
Register the Local M MCP server in PolicyLayer and add a rule for video_index: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Local Mmcp. Nothing to install.
video_index is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the video_index rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for video_index. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
video_index is provided by the Local M MCP server (rorojiao/local-mmcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →