Computes per-feature range, mean, standard deviation, percentiles, and NaN ratios, then flags anomalous windows with robust z-scores to validate embodied-AI datasets.
Compute per-feature statistics (range, mean, std, percentiles, NaN ratio) and detect anomalous windows (jumps, stalls, non-finite spans) with robust z-scoring.
| Name | Required | Description | Default |
|---|---|---|---|
| z | No | ||
| path | Yes | ||
| episodes | No | ||
| features | No | ||
| skip_anomalies | No | ||
| include_segments | No |
Changes observed during successful MCP inspections.
v0.1.0Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden. It does disclose the analysis method ('robust z-scoring') and the anomaly classes detected, which is real added context, but it never says whether the tool is read-only, whether it writes a report to disk, or whether results are returned inline, which matters for a tool that may be run against large datasets.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single dense sentence that front-loads the two main capabilities and contains no filler. Every clause (the statistic list, the anomaly types, the z-scoring method) contributes distinct information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a six-parameter, mutation-status-unknown tool with no annotations, no output schema, and zero parameter documentation, the description leaves too much open. It does not say what the return value looks like (statistics table? anomaly list?) or how the boolean flags change output, which the agent needs in order to call it correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0% across six parameters, so the description must compensate and largely does not. 'Robust z-scoring' hints at the meaning of the z threshold parameter, but features, episodes, skip_anomalies, include_segments, and the required path are entirely unexplained, so an agent cannot infer what skip_anomalies does versus simply reading the anomaly output.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description names two specific operations ('compute per-feature statistics' and 'detect anomalous windows') with concrete outputs (range, mean, std, percentiles, NaN ratio; jumps, stalls, non-finite spans). It is clearly a diagnostic tool, distinguishable in spirit from summarize_dataset or validate_dataset, but it never explicitly names what resource 'path' points to or contrasts itself against those siblings.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
There is no statement of when to choose this over summarize_dataset, validate_dataset, or detect_dataset, nor any preconditions or exclusions. The only usage signal is implicit in the word 'quality', leaving the agent to guess whether this is a pre-flight check, a profiling step, or an anomaly audit.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.