io.github.diegoperezg7/evalsmith-mcp
NPM · EVALSMITH-MCP
Evaluate RAG retrieval, agent tool calls and prompt regressions with deterministic metrics.
Available components
Last scanned (8 hours ago)
Scores higher than 41% of npm servers
Category Testing and QA
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- No install/post-install scripts declared.Pass
- 34 of 107 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency48
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 7 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability81
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (good).Pass
- Tool/resource definitions use about 1308 tokens (~72/item across 18 items; 5 tools + 13 resources), lean.Pass
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management0
- Stability not yet verified: not enough scan history yet (needs a 30-day window).Unverified
Tool Coverage97
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 91% of tool parameters carry a description.Partial
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 5 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 7 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Unverified: 1 category
A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 29 Oct 26 +15
- Malware scan: unverified → pass ▲ security
- 28 Oct 26 56
First indexed and scored.
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 5 Nov 2026 · Analysed npm/evalsmith-mcp@0.1.0
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | npm |
Background: How many MCP packages publish verified provenance →
Dependencies 107 packages
| Packages resolved | 107 |
|---|---|
| Stale | 34 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
check_tool_calls Check agent tool calls ~157
Compare an agent's tool-call trace with the expected calls: right tool, right order (strict or lax) and arguments that satisfy exact / regex / subset / any matchers. Returns a pass flag, a score and a reason for every mismatch.
| Name | Type | Req | Description |
|---|---|---|---|
| expected | array | yes | Expected calls. Argument matchers: {"type":"exact","value":...}, {"type":"regex","pattern":"..."}, {"type":"subset","value":{...}}, {"type":"any"}. |
| order | string | – | strict: the trace must be exactly the expected calls, in order. lax: every expected call must appear somewhere; order and extra calls are allowed. |
| trace | array | yes | Tool calls the agent made, in order. |
| Name | Type | Req | Description |
|---|---|---|---|
| order | string | yes | strict: the trace must be exactly the expected calls, in order. lax: every expected call must appear somewhere; order and extra calls are allowed. |
| passed | boolean | yes | – |
| score | number | yes | Matched expected calls / expected calls. |
| steps | array | yes | – |
| summary | string | yes | – |
| unexpected | array | yes | – |
No examples provided.
compare_configs Compare retrieval configurations ~218
Compare 2-3 retrieval configurations on the same qrels. Ranks them on a primary metric, attaches seeded paired-bootstrap confidence intervals, and says whether the winner's lead over the runner-up is significant (interval of the gap excludes zero).
| Name | Type | Req | Description |
|---|---|---|---|
| bootstrap | object | – | Paired bootstrap over queries. Defaults: 1000 samples, seed 42, 95% confidence. |
| configs | array | yes | Two or three runs over the same queries. |
| k | array | – | Cut-offs to evaluate. Defaults to [1, 3, 5, 10]. |
| metric | string | – | Primary metric that decides the winner, e.g. ndcg@10, recall@5, mrr. |
| qrels | object | yes | Relevance judgements: { queryId: { docId: grade } }. Grade 0 means not relevant; higher grades are more relevant (used by nDCG). Example: { "q1": { "d3": 2, "d7": 1 } } |
| Name | Type | Req | Description |
|---|---|---|---|
| bootstrap | object | yes | – |
| configs | array | yes | – |
| delta | object | yes | Paired bootstrap interval of winner minus runner-up. |
| metric | string | yes | – |
| runnerUp | string | yes | – |
| significant | boolean | yes | True when the delta interval excludes zero. |
| summary | string | yes | – |
| winner | string | yes | – |
No examples provided.
detect_prompt_regression Detect prompt regression ~157
Run a golden set's deterministic checks (exact, regex, contains, JSON Schema, length) against baseline and candidate outputs. A regression is a case that passes on baseline and fails on candidate. An LLM judge can be enabled per call for a second opinion; the verdict only depends on the checks.
| Name | Type | Req | Description |
|---|---|---|---|
| baseline | object | yes | Outputs of the current prompt/model, keyed by case id. |
| candidate | object | yes | Outputs of the proposed prompt/model, keyed by case id. |
| golden | array | yes | Golden cases with deterministic checks. |
| judge | object | – | Optional pairwise LLM judge on the server's configured OpenAI-compatible endpoint (e.g. Ollama). Off by default; it never changes the verdict. |
| Name | Type | Req | Description |
|---|---|---|---|
| baselinePassRate | number | yes | – |
| candidatePassRate | number | yes | – |
| cases | integer | yes | – |
| improvements | array | yes | Cases that fail on baseline and pass on candidate. |
| judge | object | – | – |
| perCase | array | yes | – |
| regressions | array | yes | Cases that pass on baseline and fail on candidate. |
| summary | string | yes | – |
| verdict | string | yes | – |
No examples provided.
evaluate_retrieval Evaluate retrieval ~204
Score a ranked retrieval run against relevance judgements (qrels). Returns recall@k, precision@k, nDCG@k (graded, linear gain), hit rate@k and MRR, averaged over judged queries, plus per-query values. Deterministic; no model calls.
| Name | Type | Req | Description |
|---|---|---|---|
| k | array | – | Cut-offs to evaluate. Defaults to [1, 3, 5, 10]. |
| qrels | object | yes | Relevance judgements: { queryId: { docId: grade } }. Grade 0 means not relevant; higher grades are more relevant (used by nDCG). Example: { "q1": { "d3": 2, "d7": 1 } } |
| run | object | yes | Ranked retrieval output: { queryId: [docId, ...] }, best match first. Example: { "q1": ["d7", "d3", "d9"] } |
| Name | Type | Req | Description |
|---|---|---|---|
| metrics | object | yes | Metric name to value, e.g. { "recall@5": 0.83, "mrr": 0.71 } |
| missingFromRun | array | yes | Judged queries absent from the run (scored as an empty ranking). |
| perQuery | array | yes | – |
| queries | integer | yes | Queries with at least one relevant document. |
| skippedQueries | array | yes | Queries in qrels with no relevant documents (excluded from averages). |
No examples provided.
generate_report Generate report ~147
Turn the results of the other evalsmith tools into a Markdown summary, a JSON document or JUnit XML for CI dashboards. Optional gates (metric -> minimum) decide the overall pass/fail.
| Name | Type | Req | Description |
|---|---|---|---|
| comparison | object | – | structuredContent of compare_configs. |
| format | string | – | – |
| gates | object | – | Minimum values, e.g. { "recall@5": 0.8, "tool_calls": 1 }. |
| name | string | – | – |
| regression | object | – | structuredContent of detect_prompt_regression. |
| retrieval | object | – | structuredContent of evaluate_retrieval. |
| toolCalls | object | – | structuredContent of check_tool_calls. |
| Name | Type | Req | Description |
|---|---|---|---|
| content | string | yes | – |
| format | string | yes | – |
| passed | boolean | yes | – |
No examples provided.
How do I install the io.github.diegoperezg7/evalsmith-mcp server?
io.github.diegoperezg7/evalsmith-mcp runs locally as an npm package, launched with npx -y evalsmith-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · evalsmith-mcp
claude mcp add diegoperezg7-evalsmith-mcp -- npx -y evalsmith-mcp
What is the io.github.diegoperezg7/evalsmith-mcp server?
io.github.diegoperezg7/evalsmith-mcp is listed in the public MCP registry as io.github.diegoperezg7/evalsmith-mcp. Evaluate RAG retrieval, agent tool calls and prompt regressions with deterministic metrics. This page covers its npm package (evalsmith-mcp).
Is the io.github.diegoperezg7/evalsmith-mcp server safe to use?
io.github.diegoperezg7/evalsmith-mcp scores 71 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 5 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the io.github.diegoperezg7/evalsmith-mcp server expose?
io.github.diegoperezg7/evalsmith-mcp exposes 5 tools: evaluate_retrieval, compare_configs, check_tool_calls, detect_prompt_regression, generate_report. Their descriptions and schemas cost roughly 883 tokens of context every time the server is loaded.
Is the io.github.diegoperezg7/evalsmith-mcp server still maintained?
io.github.diegoperezg7/evalsmith-mcp is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the io.github.diegoperezg7/evalsmith-mcp server under?
io.github.diegoperezg7/evalsmith-mcp declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.