@langwatch/mcp-server 1.0.0 → 2.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (204) hide show
  1. package/.python-version +1 -0
  2. package/CHANGELOG.md +110 -0
  3. package/CONTRIBUTING.md +1 -1
  4. package/README.md +86 -4
  5. package/dist/{archive-project-MK3673A7.js → archive-project-W3IKI7UL.js} +5 -5
  6. package/dist/archive-run-plan-474OKHHU.js +16 -0
  7. package/dist/{archive-scenario-KATSXRIN.js → archive-scenario-V5UNYK26.js} +5 -5
  8. package/dist/archive-test-suite-O7FLDMXK.js +16 -0
  9. package/dist/{assign-prompt-tag-E3C5W6PI.js → assign-prompt-tag-Q46WRB6P.js} +4 -4
  10. package/dist/chunk-56OGTS24.js +44 -0
  11. package/dist/chunk-6DJDEXET.js +63 -0
  12. package/dist/chunk-7E4FZIYJ.js +15 -0
  13. package/dist/{chunk-J45SVWPR.js → chunk-AN4OUA4D.js} +2 -2
  14. package/dist/chunk-CKNAH5TG.js +74 -0
  15. package/dist/{chunk-CE6QMZOP.js → chunk-EYXMLE5X.js} +86 -11
  16. package/dist/{chunk-XY4STOBE.js → chunk-G4ZAZTZW.js} +2 -2
  17. package/dist/{chunk-5UWY24RP.js → chunk-GELJ2MQR.js} +12 -3
  18. package/dist/chunk-HV4DWKBI.js +46 -0
  19. package/dist/chunk-JR6DJZOU.js +49 -0
  20. package/dist/{chunk-DAIQKS27.js → chunk-JTQJRNFJ.js} +2 -2
  21. package/dist/{chunk-WWMDHZZV.js → chunk-K37ZNPXK.js} +2 -2
  22. package/dist/{chunk-EU22L52K.js → chunk-OJLYYJWI.js} +2 -2
  23. package/dist/{chunk-Y3HDZ6OW.js → chunk-QAMP3CGR.js} +2 -2
  24. package/dist/{chunk-VYPSKRPN.js → chunk-QRO35SOM.js} +347 -146
  25. package/dist/{chunk-O3LWVN67.js → chunk-RVAE2MEA.js} +2 -2
  26. package/dist/{chunk-UXXA4D6K.js → chunk-S4GVX36W.js} +2 -2
  27. package/dist/{chunk-BJ7B4R26.js → chunk-SAS5OB7F.js} +18 -2
  28. package/dist/{chunk-RYRB2YVJ.js → chunk-SR6QPRIE.js} +2 -2
  29. package/dist/chunk-UF6XIGYR.js +58 -0
  30. package/dist/{chunk-T7WJVS2N.js → chunk-UFOYFYI7.js} +2 -2
  31. package/dist/chunk-UJPXFA26.js +267 -0
  32. package/dist/{chunk-KOO5733S.js → chunk-WT4PO4A3.js} +8 -2
  33. package/dist/config.js +2 -2
  34. package/dist/{create-agent-7W57J7MT.js → create-agent-JFJ2K6X3.js} +5 -5
  35. package/dist/{create-annotation-W2XELBCQ.js → create-annotation-EUXXQ6NS.js} +5 -5
  36. package/dist/{create-api-key-MKRXKFJM.js → create-api-key-BGBB76JX.js} +5 -5
  37. package/dist/{create-dashboard-QGL7IODC.js → create-dashboard-Z76AIPLS.js} +5 -5
  38. package/dist/{create-dataset-3PH7A4HG.js → create-dataset-FD5VAXU4.js} +5 -5
  39. package/dist/{create-dataset-records-4DZS3VRL.js → create-dataset-records-WQUXXOHI.js} +5 -5
  40. package/dist/{create-evaluator-YWH6NC4M.js → create-evaluator-6NV33Z52.js} +5 -5
  41. package/dist/create-mcp-server.js +9 -4
  42. package/dist/{create-project-BOBQ32C5.js → create-project-ZQ7TFIO2.js} +5 -5
  43. package/dist/{create-prompt-FLCFGITE.js → create-prompt-BS6QRCUU.js} +4 -4
  44. package/dist/{create-prompt-tag-EMPNIKEA.js → create-prompt-tag-FCQ5ELSQ.js} +4 -4
  45. package/dist/{create-scenario-BSBIHUL6.js → create-scenario-DRPTPXU3.js} +12 -5
  46. package/dist/create-test-suite-ZIZCCFJ4.js +37 -0
  47. package/dist/{delete-agent-7MSYNY5E.js → delete-agent-SA4JMUZF.js} +5 -5
  48. package/dist/{delete-annotation-ZT5IGN24.js → delete-annotation-X3KKK7QY.js} +5 -5
  49. package/dist/{delete-dashboard-BTIY5S3L.js → delete-dashboard-LYMRPF27.js} +5 -5
  50. package/dist/{delete-dataset-U45HQHPR.js → delete-dataset-W6EZ44HY.js} +5 -5
  51. package/dist/{delete-dataset-records-R43IMAWB.js → delete-dataset-records-7CNOTALB.js} +5 -5
  52. package/dist/{delete-evaluator-U6JMQ6C7.js → delete-evaluator-RKYC4TNB.js} +5 -5
  53. package/dist/{delete-prompt-tag-4EV4TDHB.js → delete-prompt-tag-QB3OCZUV.js} +4 -4
  54. package/dist/{delete-workflow-5QYALXEL.js → delete-workflow-NG2VXYYM.js} +5 -5
  55. package/dist/{discover-evaluator-schema-MRGHVNTT.js → discover-evaluator-schema-ESSHQGPY.js} +58 -321
  56. package/dist/{discover-scenario-schema-UPBTCWPG.js → discover-scenario-schema-W7KBXA4J.js} +54 -2
  57. package/dist/{discover-schema-ICGXPHUW.js → discover-schema-L6QDPON3.js} +2 -2
  58. package/dist/get-agent-7THF7HCU.js +66 -0
  59. package/dist/{get-analytics-PVERWOA5.js → get-analytics-R6AZD3AW.js} +4 -4
  60. package/dist/{get-dashboard-ZZPC3WMD.js → get-dashboard-P4WWVHFU.js} +5 -5
  61. package/dist/{get-dataset-25TB2LQH.js → get-dataset-AKTYSQDI.js} +5 -5
  62. package/dist/{get-evaluator-3FTRQKEC.js → get-evaluator-RZ44UIN6.js} +5 -5
  63. package/dist/{get-project-5OAXYC57.js → get-project-QSQEAB33.js} +5 -5
  64. package/dist/get-prompt-GVKXF5NL.js +100 -0
  65. package/dist/get-run-plan-NY7J5GRE.js +61 -0
  66. package/dist/{get-scenario-DSSTR6FI.js → get-scenario-Q3LLQE4T.js} +9 -5
  67. package/dist/{get-simulation-run-RQPMS2MC.js → get-simulation-run-WINEYDMK.js} +14 -5
  68. package/dist/get-test-suite-4HKA7OKS.js +49 -0
  69. package/dist/{get-trace-O5VT3NAC.js → get-trace-RCUV7N4O.js} +4 -4
  70. package/dist/{get-workflow-3EZDOEIF.js → get-workflow-EEOC7Y45.js} +5 -5
  71. package/dist/http-security-EU3CHTY7.js +26 -0
  72. package/dist/{http-server-2ACJ3CXW.js → http-server-WCDUARTH.js} +1038 -693
  73. package/dist/index.js +6437 -4020
  74. package/dist/{langwatch-api-annotations-E65MI3WZ.js → langwatch-api-annotations-6DHVZWUK.js} +5 -5
  75. package/dist/{langwatch-api-dashboards-QCDECRCY.js → langwatch-api-dashboards-2KJL2M42.js} +5 -5
  76. package/dist/{langwatch-api-monitors-UIO2KTDV.js → langwatch-api-monitors-6ISFJ7WX.js} +4 -4
  77. package/dist/{langwatch-api-secrets-4T2EKSJ7.js → langwatch-api-secrets-OEBD6GBU.js} +4 -4
  78. package/dist/{langwatch-api-triggers-MPOEKPUY.js → langwatch-api-triggers-5E6DDGX6.js} +4 -4
  79. package/dist/langwatch-api-workflows-6UJL5U4Z.js +16 -0
  80. package/dist/list-agents-AMVDQPQL.js +43 -0
  81. package/dist/{list-annotations-CGX5BT7E.js → list-annotations-6PR5NLIC.js} +5 -5
  82. package/dist/{list-api-keys-JLWJPFRX.js → list-api-keys-GVPNBIHQ.js} +5 -5
  83. package/dist/{list-dashboards-RNIWX3TX.js → list-dashboards-Z3BZKDGF.js} +5 -5
  84. package/dist/{list-dataset-records-B5MHZR66.js → list-dataset-records-O67HZ35Y.js} +5 -5
  85. package/dist/{list-datasets-BVCAMQCP.js → list-datasets-WXZL6UMP.js} +5 -5
  86. package/dist/{list-evaluators-65FBTL2U.js → list-evaluators-PAX72WCN.js} +5 -5
  87. package/dist/{list-model-providers-GRK3I3VV.js → list-model-providers-VVLZMLCX.js} +5 -5
  88. package/dist/{list-projects-JLMBFMW3.js → list-projects-AQVI6SUK.js} +5 -5
  89. package/dist/{list-prompt-tags-UHNWDGQI.js → list-prompt-tags-WZFBJOHG.js} +4 -4
  90. package/dist/{list-prompts-7254B3U7.js → list-prompts-BQOHRHYT.js} +4 -4
  91. package/dist/list-run-plans-UBAU2DGH.js +46 -0
  92. package/dist/{list-scenarios-46YNICKM.js → list-scenarios-MS4IUIVM.js} +10 -7
  93. package/dist/{list-simulation-runs-LGNEFNPN.js → list-simulation-runs-D6YF4ZOU.js} +13 -7
  94. package/dist/list-test-suites-L7R2O6DR.js +35 -0
  95. package/dist/{list-workflows-MLQDM6CP.js → list-workflows-6MV4XJHF.js} +5 -5
  96. package/dist/{rename-prompt-tag-RHQMNEQL.js → rename-prompt-tag-K5LQ4CY6.js} +4 -4
  97. package/dist/rename-test-suite-NTYHAO5X.js +21 -0
  98. package/dist/report-issue-RHZZ33IS.js +885 -0
  99. package/dist/rerun-run-plan-MUDWJBE4.js +22 -0
  100. package/dist/{revoke-api-key-3Y35EMBT.js → revoke-api-key-6LAZ3QZZ.js} +5 -5
  101. package/dist/run-agent-CAGWGTYT.js +72 -0
  102. package/dist/run-plan-T4ZYOZGA.js +39 -0
  103. package/dist/run-test-suite-L4MZ7G4D.js +26 -0
  104. package/dist/{run-workflow-5LVBZYLS.js → run-workflow-DFIS23AP.js} +5 -5
  105. package/dist/{search-traces-U4HLOCZ2.js → search-traces-4XTJYXIK.js} +4 -4
  106. package/dist/{set-model-provider-V3M5CDIY.js → set-model-provider-XRNMFOVR.js} +5 -5
  107. package/dist/{update-agent-47RZIHS7.js → update-agent-DUWNYBNA.js} +5 -5
  108. package/dist/{update-dataset-G6QJJGS5.js → update-dataset-YSOMDCOB.js} +5 -5
  109. package/dist/{update-dataset-record-RMWLSWCS.js → update-dataset-record-DEA2ITYR.js} +5 -5
  110. package/dist/{update-evaluator-SFRRI5Q3.js → update-evaluator-NH7Q5GWV.js} +5 -5
  111. package/dist/{update-project-PI5DWBVK.js → update-project-Q26FBSYH.js} +5 -5
  112. package/dist/update-prompt-ZQQ2C3KU.js +143 -0
  113. package/dist/{update-scenario-G3IU2UIB.js → update-scenario-WUX34JAR.js} +12 -5
  114. package/package.json +8 -9
  115. package/pyproject.toml +5 -5
  116. package/src/__tests__/agent-tools.unit.test.ts +257 -0
  117. package/src/__tests__/all-tools.integration.test.ts +321 -75
  118. package/src/__tests__/config.unit.test.ts +33 -0
  119. package/src/__tests__/docker.integration.test.ts +91 -7
  120. package/src/__tests__/documentation-fetch-security.integration.test.ts +8 -2
  121. package/src/__tests__/feature-map-tool-drift.unit.test.ts +93 -0
  122. package/src/__tests__/http-security-origin.unit.test.ts +97 -0
  123. package/src/__tests__/http-security-sessions.unit.test.ts +209 -0
  124. package/src/__tests__/http-security-verifier.unit.test.ts +248 -0
  125. package/src/__tests__/http-server-allocation.integration.test.ts +279 -0
  126. package/src/__tests__/http-server-origin.integration.test.ts +169 -0
  127. package/src/__tests__/http-server-session-auth.integration.test.ts +341 -0
  128. package/src/__tests__/http-transport.integration.test.ts +85 -87
  129. package/src/__tests__/integration.integration.test.ts +7 -11
  130. package/src/__tests__/prompt-response-types.unit.test.ts +51 -0
  131. package/src/__tests__/prompts.integration.test.ts +736 -0
  132. package/src/__tests__/run-plan-tools.unit.test.ts +565 -0
  133. package/src/__tests__/scenario-tools.unit.test.ts +19 -0
  134. package/src/__tests__/simulation-run-tools.unit.test.ts +116 -0
  135. package/src/__tests__/support/http-server-harness.ts +138 -0
  136. package/src/__tests__/test-suite-tools.unit.test.ts +615 -0
  137. package/src/__tests__/{tools.unit.test.ts → tools.integration.test.ts} +31 -59
  138. package/src/config.ts +17 -2
  139. package/src/create-mcp-server.ts +334 -55
  140. package/src/http-security.ts +492 -0
  141. package/src/http-server.ts +536 -152
  142. package/src/index.ts +49 -2
  143. package/src/langwatch-api-agents.ts +196 -12
  144. package/src/langwatch-api-run-plans.ts +157 -0
  145. package/src/langwatch-api-scenarios.ts +9 -0
  146. package/src/langwatch-api-simulation-runs.ts +5 -0
  147. package/src/langwatch-api-test-suites.ts +135 -0
  148. package/src/langwatch-api.ts +62 -12
  149. package/src/schemas/index.ts +5 -0
  150. package/src/schemas/run-plan.ts +92 -0
  151. package/src/schemas/suite-fields.ts +149 -0
  152. package/src/tools/archive-run-plan.ts +12 -0
  153. package/src/tools/archive-test-suite.ts +12 -0
  154. package/src/tools/create-scenario.ts +8 -0
  155. package/src/tools/create-test-suite.ts +43 -0
  156. package/src/tools/discover-evaluator-schema.ts +2 -2
  157. package/src/tools/discover-scenario-schema.ts +54 -0
  158. package/src/tools/format-run-plan.ts +68 -0
  159. package/src/tools/format-scenario.ts +14 -0
  160. package/src/tools/format-suite-details.ts +96 -0
  161. package/src/tools/get-agent.ts +45 -3
  162. package/src/tools/get-prompt.ts +108 -24
  163. package/src/tools/get-run-plan.ts +62 -0
  164. package/src/tools/get-scenario.ts +3 -0
  165. package/src/tools/get-simulation-run.ts +9 -0
  166. package/src/tools/get-test-suite.ts +49 -0
  167. package/src/tools/list-agents.ts +20 -3
  168. package/src/tools/list-run-plans.ts +47 -0
  169. package/src/tools/list-scenarios.ts +12 -2
  170. package/src/tools/list-simulation-runs.ts +6 -2
  171. package/src/tools/list-test-suites.ts +35 -0
  172. package/src/tools/rename-test-suite.ts +18 -0
  173. package/src/tools/rerun-run-plan.ts +21 -0
  174. package/src/tools/run-agent.ts +54 -6
  175. package/src/tools/run-plan.ts +49 -0
  176. package/src/tools/run-test-suite.ts +30 -0
  177. package/src/tools/test-agent.ts +18 -0
  178. package/src/tools/update-prompt.ts +195 -10
  179. package/src/tools/update-scenario.ts +8 -0
  180. package/src/tools/update-test-suite.ts +45 -0
  181. package/tsconfig.json +7 -2
  182. package/uv.lock +418 -298
  183. package/dist/archive-suite-C3KX6IIZ.js +0 -16
  184. package/dist/chunk-WDH5P5RF.js +0 -48
  185. package/dist/create-suite-TWF6BBR4.js +0 -39
  186. package/dist/get-agent-7QJVPH72.js +0 -29
  187. package/dist/get-prompt-I753LZKZ.js +0 -46
  188. package/dist/get-suite-GTOCY5PL.js +0 -45
  189. package/dist/langwatch-api-workflows-5HRZ7HR4.js +0 -16
  190. package/dist/list-agents-TIJTP566.js +0 -33
  191. package/dist/list-suites-5VK4HXEK.js +0 -43
  192. package/dist/report-issue-DXJ6TPTC.js +0 -412
  193. package/dist/run-agent-RBSIFKRT.js +0 -37
  194. package/dist/run-suite-K3DNQ3SE.js +0 -30
  195. package/dist/update-prompt-RQFTEO7Y.js +0 -25
  196. package/dist/update-suite-TRGGLNR7.js +0 -40
  197. package/pnpm-workspace.yaml +0 -34
  198. package/src/langwatch-api-suites.ts +0 -118
  199. package/src/tools/archive-suite.ts +0 -12
  200. package/src/tools/create-suite.ts +0 -42
  201. package/src/tools/get-suite.ts +0 -47
  202. package/src/tools/list-suites.ts +0 -43
  203. package/src/tools/run-suite.ts +0 -28
  204. package/src/tools/update-suite.ts +0 -44
@@ -0,0 +1,138 @@
1
+ /**
2
+ * Shared fixtures for the standalone MCP HTTP server tests.
3
+ */
4
+
5
+ import { expect, vi } from "vitest";
6
+
7
+ import type { ApiKeyVerifier } from "../../http-security.js";
8
+ import { startHttpServer } from "../../http-server.js";
9
+
10
+ /** Standard headers the MCP Streamable HTTP protocol requires on POST. */
11
+ export const MCP_POST_HEADERS = {
12
+ "Content-Type": "application/json",
13
+ Accept: "application/json, text/event-stream",
14
+ };
15
+
16
+ export const VALID_KEY = "sk-lw-valid-key";
17
+ export const OTHER_VALID_KEY = "sk-lw-other-valid-key";
18
+
19
+ /**
20
+ * Stands in for the LangWatch API. The standalone server has no database, so
21
+ * it asks the API whether a key is real; here a fixed list is.
22
+ */
23
+ export function countingVerifier(validKeys: string[]) {
24
+ const verify = vi.fn(async (apiKey: string) => validKeys.includes(apiKey));
25
+ const verifier: ApiKeyVerifier = {
26
+ verify,
27
+ sweep: () => undefined,
28
+ clear: () => undefined,
29
+ };
30
+ return { verifier, verify };
31
+ }
32
+
33
+ /** A verifier whose accepted keys can change mid-test, to model revocation. */
34
+ export function revocableVerifier(initialKeys: string[]) {
35
+ const live = new Set(initialKeys);
36
+ const verifier: ApiKeyVerifier = {
37
+ verify: async (apiKey: string) => live.has(apiKey),
38
+ sweep: () => undefined,
39
+ clear: () => undefined,
40
+ };
41
+ return { verifier, revoke: (apiKey: string) => live.delete(apiKey) };
42
+ }
43
+
44
+ export interface Harness {
45
+ baseUrl: string;
46
+ close: () => Promise<void>;
47
+ }
48
+
49
+ export async function startHarness(
50
+ options: Parameters<typeof startHttpServer>[0]
51
+ ): Promise<Harness> {
52
+ const { server, port, host } = await startHttpServer(options);
53
+ return {
54
+ baseUrl: `http://${host === "0.0.0.0" ? "127.0.0.1" : host}:${port}`,
55
+ close: () =>
56
+ new Promise<void>((resolve) => {
57
+ server.close(() => resolve());
58
+ }),
59
+ };
60
+ }
61
+
62
+ export function initializeBody(): string {
63
+ return JSON.stringify({
64
+ jsonrpc: "2.0",
65
+ method: "initialize",
66
+ params: {
67
+ protocolVersion: "2025-03-26",
68
+ capabilities: {},
69
+ clientInfo: { name: "test-client", version: "1.0.0" },
70
+ },
71
+ id: 1,
72
+ });
73
+ }
74
+
75
+ export function toolsListBody(): string {
76
+ return JSON.stringify({ jsonrpc: "2.0", method: "tools/list", id: 2 });
77
+ }
78
+
79
+ /** Opens a Streamable HTTP session and returns its id. */
80
+ export async function openSession({
81
+ baseUrl,
82
+ apiKey = VALID_KEY,
83
+ }: {
84
+ baseUrl: string;
85
+ apiKey?: string;
86
+ }): Promise<string> {
87
+ const response = await fetch(`${baseUrl}/mcp`, {
88
+ method: "POST",
89
+ headers: { ...MCP_POST_HEADERS, Authorization: `Bearer ${apiKey}` },
90
+ body: initializeBody(),
91
+ });
92
+ expect(response.status).toBe(200);
93
+ await response.text();
94
+ return response.headers.get("mcp-session-id")!;
95
+ }
96
+
97
+ /**
98
+ * Reads an SSE stream until the endpoint event carrying the session id has
99
+ * arrived. A single read can return a partial frame, so waiting for the
100
+ * pattern rather than the first chunk keeps this from failing intermittently.
101
+ */
102
+ export async function readSessionId(response: Response): Promise<string> {
103
+ const reader = response.body!.getReader();
104
+ const decoder = new TextDecoder();
105
+ let buffered = "";
106
+
107
+ for (let reads = 0; reads < 20; reads++) {
108
+ const { value, done } = await reader.read();
109
+ if (done) break;
110
+ buffered += decoder.decode(value, { stream: true });
111
+ const match = /sessionId=([\w-]+)/.exec(buffered);
112
+ if (match?.[1]) return match[1];
113
+ }
114
+
115
+ throw new Error(`No session id in SSE stream. Received: ${buffered}`);
116
+ }
117
+
118
+ /**
119
+ * Sends `request` repeatedly until it answers 429, and returns that response
120
+ * along with how many attempts it took. Returns `rejected: undefined` when the
121
+ * limit never triggered within `maxAttempts`.
122
+ */
123
+ export async function requestUntilThrottled({
124
+ request,
125
+ maxAttempts = 30,
126
+ }: {
127
+ request: (attempt: number) => Promise<Response>;
128
+ maxAttempts?: number;
129
+ }): Promise<{ attempts: number; rejected: Response | undefined }> {
130
+ for (let attempt = 0; attempt < maxAttempts; attempt++) {
131
+ const response = await request(attempt);
132
+ if (response.status === 429) {
133
+ return { attempts: attempt + 1, rejected: response };
134
+ }
135
+ await response.text();
136
+ }
137
+ return { attempts: maxAttempts, rejected: undefined };
138
+ }