SkillTotal
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| SKILLTOTAL_WORKERS | No | Processes a large scan uses (1 for one) | up to 4 |
| SKILLTOTAL_MAX_CLONE_MB | No | Largest git working tree cloned | 200 |
| SKILLTOTAL_CLONE_TIMEOUT | No | Seconds a git clone may take | 300 |
| SKILLTOTAL_MAX_ARCHIVE_MB | No | Largest archive downloaded or read | 150 |
| SKILLTOTAL_MAX_EXTRACT_MB | No | Largest total size an archive may extract to | 400 |
| SKILLTOTAL_MAX_READABLE_MB | No | Most text a scan reads | same as clone |
| SKILLTOTAL_MAX_ARCHIVE_MEMBERS | No | Most entries an archive may hold | 30000 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| scan_componentA | Statically scan an AI component (agent skill, MCP server, npm/PyPI package, repository, local path or archive) for supply-chain risks, dangerous capabilities, prompt-injection surfaces, and exfiltration paths. Use this BEFORE installing or trusting a component. The scan is local, deterministic, and never executes the component's code. Returns the full SkillTotal report as JSON: verdict, risk_score (0-100), findings with file/line evidence, and capabilities. |
| diff_componentsA | Compare two versions of a component and report what changed: new/resolved/changed findings, capability changes, and the risk-score delta. Use for upgrade reviews (is the new version riskier than the one already vetted?). Both sides accept the same sources as scan_component. |
| list_rulesA | List every SkillTotal detection rule: id, severity, category, title, threat class and capability. Pass verbose=true for each rule's description and recommendation. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 3 tools
The three tools target clearly distinct actions: a single-component scan, a two-version comparison, and rule introspection. There is no plausible confusion about which tool to pick for a given need, and the descriptions reinforce the boundaries (diff explicitly builds on scan's sources).
All three names follow a clean verb_noun pattern (scan_component, diff_components, list_rules) in snake_case with consistent pluralization conventions. An agent can reliably infer the tool's purpose from its name alone.
Three tools is lean but well-matched to a focused static-analysis server: scan, compare, and inspect rules form a coherent minimal surface. It is slightly under-provisioned for things like batch scanning or report history, but nothing is gratuitous.
The core supply-chain vetting workflow is covered end-to-end: scan a component, diff two versions for upgrade review, and enumerate detection rules to interpret findings. Minor gaps exist around retrieving prior reports or scanning many components at once, but agents can work around these by re-scanning.