@hoilab/ada-cli 0.84.21 → 0.84.22

package/dist/utils/frontmatter.js

@@ -1,7 +1,8 @@
 import { parse } from "yaml";
+import { stripBom } from "./text.js";
 const normalizeNewlines = (value) => value.replace(/\r\n/g, "\n").replace(/\r/g, "\n");
 const extractFrontmatter = (content) => {
-    const normalized = normalizeNewlines(content);
+    const normalized = normalizeNewlines(stripBom(content));
     if (!normalized.startsWith("---")) {
         return { yamlString: null, body: normalized };
     }

package/dist/utils/frontmatter.js.map

@@ -1 +1 @@
-{"version":3,"file":"frontmatter.js","sourceRoot":"","sources":["../../src/utils/frontmatter.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,MAAM,CAAC;AAO7B,MAAM,iBAAiB,GAAG,CAAC,KAAa,EAAU,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,IAAI,CAAC,CAAC;AAEvG,MAAM,kBAAkB,GAAG,CAAC,OAAe,EAA+C,EAAE,CAAC;IAC5F,MAAM,UAAU,GAAG,iBAAiB,CAAC,OAAO,CAAC,CAAC;IAE9C,IAAI,CAAC,UAAU,CAAC,UAAU,CAAC,KAAK,CAAC,EAAE,CAAC;QACnC,OAAO,EAAE,UAAU,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC;IAC/C,CAAC;IAED,MAAM,QAAQ,GAAG,UAAU,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;IAChD,IAAI,QAAQ,KAAK,CAAC,CAAC,EAAE,CAAC;QACrB,OAAO,EAAE,UAAU,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC;IAC/C,CAAC;IAED,OAAO;QACN,UAAU,EAAE,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,CAAC;QACzC,IAAI,EAAE,UAAU,CAAC,KAAK,CAAC,QAAQ,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE;KAC3C,CAAC;AAAA,CACF,CAAC;AAEF,MAAM,CAAC,MAAM,gBAAgB,GAAG,CAC/B,OAAe,EACQ,EAAE,CAAC;IAC1B,MAAM,EAAE,UAAU,EAAE,IAAI,EAAE,GAAG,kBAAkB,CAAC,OAAO,CAAC,CAAC;IACzD,IAAI,CAAC,UAAU,EAAE,CAAC;QACjB,OAAO,EAAE,WAAW,EAAE,EAAO,EAAE,IAAI,EAAE,CAAC;IACvC,CAAC;IACD,MAAM,MAAM,GAAG,KAAK,CAAC,UAAU,CAAC,CAAC;IACjC,OAAO,EAAE,WAAW,EAAE,CAAC,MAAM,IAAI,EAAE,CAAM,EAAE,IAAI,EAAE,CAAC;AAAA,CAClD,CAAC;AAEF,MAAM,CAAC,MAAM,gBAAgB,GAAG,CAAC,OAAe,EAAU,EAAE,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC","sourcesContent":["import { parse } from \"yaml\";\n\ntype ParsedFrontmatter<T extends Record<string, unknown>> = {\n\tfrontmatter: T;\n\tbody: string;\n};\n\nconst normalizeNewlines = (value: string): string => value.replace(/\\r\\n/g, \"\\n\").replace(/\\r/g, \"\\n\");\n\nconst extractFrontmatter = (content: string): { yamlString: string | null; body: string } => {\n\tconst normalized = normalizeNewlines(content);\n\n\tif (!normalized.startsWith(\"---\")) {\n\t\treturn { yamlString: null, body: normalized };\n\t}\n\n\tconst endIndex = normalized.indexOf(\"\\n---\", 3);\n\tif (endIndex === -1) {\n\t\treturn { yamlString: null, body: normalized };\n\t}\n\n\treturn {\n\t\tyamlString: normalized.slice(4, endIndex),\n\t\tbody: normalized.slice(endIndex + 4).trim(),\n\t};\n};\n\nexport const parseFrontmatter = <T extends Record<string, unknown> = Record<string, unknown>>(\n\tcontent: string,\n): ParsedFrontmatter<T> => {\n\tconst { yamlString, body } = extractFrontmatter(content);\n\tif (!yamlString) {\n\t\treturn { frontmatter: {} as T, body };\n\t}\n\tconst parsed = parse(yamlString);\n\treturn { frontmatter: (parsed ?? {}) as T, body };\n};\n\nexport const stripFrontmatter = (content: string): string => parseFrontmatter(content).body;\n"]}
+{"version":3,"file":"frontmatter.js","sourceRoot":"","sources":["../../src/utils/frontmatter.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,MAAM,CAAC;AAC7B,OAAO,EAAE,QAAQ,EAAE,MAAM,WAAW,CAAC;AAOrC,MAAM,iBAAiB,GAAG,CAAC,KAAa,EAAU,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,IAAI,CAAC,CAAC;AAEvG,MAAM,kBAAkB,GAAG,CAAC,OAAe,EAA+C,EAAE,CAAC;IAC5F,MAAM,UAAU,GAAG,iBAAiB,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC;IAExD,IAAI,CAAC,UAAU,CAAC,UAAU,CAAC,KAAK,CAAC,EAAE,CAAC;QACnC,OAAO,EAAE,UAAU,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC;IAC/C,CAAC;IAED,MAAM,QAAQ,GAAG,UAAU,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;IAChD,IAAI,QAAQ,KAAK,CAAC,CAAC,EAAE,CAAC;QACrB,OAAO,EAAE,UAAU,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC;IAC/C,CAAC;IAED,OAAO;QACN,UAAU,EAAE,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,CAAC;QACzC,IAAI,EAAE,UAAU,CAAC,KAAK,CAAC,QAAQ,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE;KAC3C,CAAC;AAAA,CACF,CAAC;AAEF,MAAM,CAAC,MAAM,gBAAgB,GAAG,CAC/B,OAAe,EACQ,EAAE,CAAC;IAC1B,MAAM,EAAE,UAAU,EAAE,IAAI,EAAE,GAAG,kBAAkB,CAAC,OAAO,CAAC,CAAC;IACzD,IAAI,CAAC,UAAU,EAAE,CAAC;QACjB,OAAO,EAAE,WAAW,EAAE,EAAO,EAAE,IAAI,EAAE,CAAC;IACvC,CAAC;IACD,MAAM,MAAM,GAAG,KAAK,CAAC,UAAU,CAAC,CAAC;IACjC,OAAO,EAAE,WAAW,EAAE,CAAC,MAAM,IAAI,EAAE,CAAM,EAAE,IAAI,EAAE,CAAC;AAAA,CAClD,CAAC;AAEF,MAAM,CAAC,MAAM,gBAAgB,GAAG,CAAC,OAAe,EAAU,EAAE,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC","sourcesContent":["import { parse } from \"yaml\";\nimport { stripBom } from \"./text.ts\";\n\ntype ParsedFrontmatter<T extends Record<string, unknown>> = {\n\tfrontmatter: T;\n\tbody: string;\n};\n\nconst normalizeNewlines = (value: string): string => value.replace(/\\r\\n/g, \"\\n\").replace(/\\r/g, \"\\n\");\n\nconst extractFrontmatter = (content: string): { yamlString: string | null; body: string } => {\n\tconst normalized = normalizeNewlines(stripBom(content));\n\n\tif (!normalized.startsWith(\"---\")) {\n\t\treturn { yamlString: null, body: normalized };\n\t}\n\n\tconst endIndex = normalized.indexOf(\"\\n---\", 3);\n\tif (endIndex === -1) {\n\t\treturn { yamlString: null, body: normalized };\n\t}\n\n\treturn {\n\t\tyamlString: normalized.slice(4, endIndex),\n\t\tbody: normalized.slice(endIndex + 4).trim(),\n\t};\n};\n\nexport const parseFrontmatter = <T extends Record<string, unknown> = Record<string, unknown>>(\n\tcontent: string,\n): ParsedFrontmatter<T> => {\n\tconst { yamlString, body } = extractFrontmatter(content);\n\tif (!yamlString) {\n\t\treturn { frontmatter: {} as T, body };\n\t}\n\tconst parsed = parse(yamlString);\n\treturn { frontmatter: (parsed ?? {}) as T, body };\n};\n\nexport const stripFrontmatter = (content: string): string => parseFrontmatter(content).body;\n"]}

package/dist/utils/management-http.d.ts

@@ -4,8 +4,10 @@ export interface FetchRetryOptions {
     maxRetries?: number;
     /** Retry transient HTTP responses as well as transport failures. Defaults to true. */
     retryOnStatus?: boolean;
-    /** Per-attempt timeout. A new timeout is created for every attempt. */
+    /** Overall time budget shared by all attempts. */
     timeoutMs?: number;
+    /** Per-attempt timeout. A new timeout is created for every attempt. */
+    attemptTimeoutMs?: number;
 }
 /**
  * Fetch a management HTTP resource with a bounded immediate retry.
@@ -15,8 +17,8 @@ export interface FetchRetryOptions {
  * agent/model operations: those can fail after the HTTP request starts and are
  * retried by their semantic caller instead.
  *
- * Caller cancellation is terminal. When timeoutMs is supplied, it is the
- * overall time budget shared by all attempts.
+ * Caller cancellation and timeoutMs are terminal. attemptTimeoutMs aborts
+ * only the current attempt so a hung connection can be retried.
  */
 export declare function fetchWithRetry(input: FetchInput, init?: RequestInit | undefined, options?: FetchRetryOptions): Promise<Response>;
 export {};

package/dist/utils/management-http.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"management-http.d.ts","sourceRoot":"","sources":["../../src/utils/management-http.ts"],"names":[],"mappings":"AAAA,KAAK,UAAU,GAAG,UAAU,CAAC,OAAO,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;AAI9C,MAAM,WAAW,iBAAiB;IACjC,gFAAgF;IAChF,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,sFAAsF;IACtF,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,uEAAuE;IACvE,SAAS,CAAC,EAAE,MAAM,CAAC;CACnB;AAED;;;;;;;;;;GAUG;AACH,wBAAsB,cAAc,CACnC,KAAK,EAAE,UAAU,EACjB,IAAI,GAAE,WAAW,GAAG,SAAqB,EACzC,OAAO,GAAE,iBAAsB,GAC7B,OAAO,CAAC,QAAQ,CAAC,CAuCnB","sourcesContent":["type FetchInput = Parameters<typeof fetch>[0];\n\nconst RETRYABLE_STATUS_CODES = new Set([408, 425, 429, 500, 502, 503, 504]);\n\nexport interface FetchRetryOptions {\n\t/** Number of additional attempts after the initial request. Defaults to two. */\n\tmaxRetries?: number;\n\t/** Retry transient HTTP responses as well as transport failures. Defaults to true. */\n\tretryOnStatus?: boolean;\n\t/** Per-attempt timeout. A new timeout is created for every attempt. */\n\ttimeoutMs?: number;\n}\n\n/**\n * Fetch a management HTTP resource with a bounded immediate retry.\n *\n * This is intentionally a transport-level helper for idempotent management\n * requests (version checks, catalogs, and downloads). It must not be used for\n * agent/model operations: those can fail after the HTTP request starts and are\n * retried by their semantic caller instead.\n *\n * Caller cancellation is terminal. When timeoutMs is supplied, it is the\n * overall time budget shared by all attempts.\n */\nexport async function fetchWithRetry(\n\tinput: FetchInput,\n\tinit: RequestInit | undefined = undefined,\n\toptions: FetchRetryOptions = {},\n): Promise<Response> {\n\tconst maxRetries =\n\t\toptions.maxRetries === undefined || !Number.isFinite(options.maxRetries)\n\t\t\t? 2\n\t\t\t: Math.max(0, Math.floor(options.maxRetries));\n\tconst retryOnStatus = options.retryOnStatus ?? true;\n\tconst parentSignal = init?.signal;\n\tconst timeoutSignal =\n\t\toptions.timeoutMs !== undefined && options.timeoutMs > 0 ? AbortSignal.timeout(options.timeoutMs) : undefined;\n\tconst signal = timeoutSignal\n\t\t? parentSignal\n\t\t\t? AbortSignal.any([parentSignal, timeoutSignal])\n\t\t\t: timeoutSignal\n\t\t: parentSignal;\n\n\tfor (let attempt = 0; ; attempt++) {\n\t\tsignal?.throwIfAborted();\n\n\t\ttry {\n\t\t\tconst response = await fetch(input, signal ? { ...init, signal } : init);\n\t\t\tconst shouldRetry = retryOnStatus && RETRYABLE_STATUS_CODES.has(response.status) && attempt < maxRetries;\n\t\t\tif (!shouldRetry) return response;\n\t\t\ttry {\n\t\t\t\tawait response.body?.cancel();\n\t\t\t} catch {\n\t\t\t\t// The response is being discarded before a retry. There is nothing useful to\n\t\t\t\t// do if cancelling its body also fails.\n\t\t\t}\n\t\t} catch (error) {\n\t\t\tif (\n\t\t\t\tparentSignal?.aborted ||\n\t\t\t\ttimeoutSignal?.aborted ||\n\t\t\t\t(error instanceof Error && error.name === \"AbortError\" && timeoutSignal === undefined) ||\n\t\t\t\tattempt >= maxRetries\n\t\t\t) {\n\t\t\t\tthrow error;\n\t\t\t}\n\t\t}\n\t}\n}\n"]}
+{"version":3,"file":"management-http.d.ts","sourceRoot":"","sources":["../../src/utils/management-http.ts"],"names":[],"mappings":"AAAA,KAAK,UAAU,GAAG,UAAU,CAAC,OAAO,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;AAI9C,MAAM,WAAW,iBAAiB;IACjC,gFAAgF;IAChF,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,sFAAsF;IACtF,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,kDAAkD;IAClD,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,uEAAuE;IACvE,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC1B;AAED;;;;;;;;;;GAUG;AACH,wBAAsB,cAAc,CACnC,KAAK,EAAE,UAAU,EACjB,IAAI,GAAE,WAAW,GAAG,SAAqB,EACzC,OAAO,GAAE,iBAAsB,GAC7B,OAAO,CAAC,QAAQ,CAAC,CA+CnB","sourcesContent":["type FetchInput = Parameters<typeof fetch>[0];\n\nconst RETRYABLE_STATUS_CODES = new Set([408, 425, 429, 500, 502, 503, 504]);\n\nexport interface FetchRetryOptions {\n\t/** Number of additional attempts after the initial request. Defaults to two. */\n\tmaxRetries?: number;\n\t/** Retry transient HTTP responses as well as transport failures. Defaults to true. */\n\tretryOnStatus?: boolean;\n\t/** Overall time budget shared by all attempts. */\n\ttimeoutMs?: number;\n\t/** Per-attempt timeout. A new timeout is created for every attempt. */\n\tattemptTimeoutMs?: number;\n}\n\n/**\n * Fetch a management HTTP resource with a bounded immediate retry.\n *\n * This is intentionally a transport-level helper for idempotent management\n * requests (version checks, catalogs, and downloads). It must not be used for\n * agent/model operations: those can fail after the HTTP request starts and are\n * retried by their semantic caller instead.\n *\n * Caller cancellation and timeoutMs are terminal. attemptTimeoutMs aborts\n * only the current attempt so a hung connection can be retried.\n */\nexport async function fetchWithRetry(\n\tinput: FetchInput,\n\tinit: RequestInit | undefined = undefined,\n\toptions: FetchRetryOptions = {},\n): Promise<Response> {\n\tconst maxRetries =\n\t\toptions.maxRetries === undefined || !Number.isFinite(options.maxRetries)\n\t\t\t? 2\n\t\t\t: Math.max(0, Math.floor(options.maxRetries));\n\tconst retryOnStatus = options.retryOnStatus ?? true;\n\tconst parentSignal = init?.signal ?? undefined;\n\tconst timeoutSignal =\n\t\toptions.timeoutMs !== undefined && options.timeoutMs > 0 ? AbortSignal.timeout(options.timeoutMs) : undefined;\n\tconst attemptTimeoutMs =\n\t\toptions.attemptTimeoutMs !== undefined && options.attemptTimeoutMs > 0 ? options.attemptTimeoutMs : undefined;\n\n\tfor (let attempt = 0; ; attempt++) {\n\t\tparentSignal?.throwIfAborted();\n\t\ttimeoutSignal?.throwIfAborted();\n\t\tconst attemptTimeoutSignal = attemptTimeoutMs ? AbortSignal.timeout(attemptTimeoutMs) : undefined;\n\t\tconst signals = [parentSignal, timeoutSignal, attemptTimeoutSignal].filter(\n\t\t\t(signal): signal is AbortSignal => signal !== undefined,\n\t\t);\n\t\tconst signal = signals.length > 1 ? AbortSignal.any(signals) : signals[0];\n\n\t\ttry {\n\t\t\tconst response = await fetch(input, signal ? { ...init, signal } : init);\n\t\t\tconst shouldRetry = retryOnStatus && RETRYABLE_STATUS_CODES.has(response.status) && attempt < maxRetries;\n\t\t\tif (!shouldRetry) return response;\n\t\t\ttry {\n\t\t\t\tawait response.body?.cancel();\n\t\t\t} catch {\n\t\t\t\t// The response is being discarded before a retry. There is nothing useful to\n\t\t\t\t// do if cancelling its body also fails.\n\t\t\t}\n\t\t} catch (error) {\n\t\t\tconst attemptTimedOut =\n\t\t\t\tattemptTimeoutSignal?.aborted === true && !parentSignal?.aborted && !timeoutSignal?.aborted;\n\t\t\tif (\n\t\t\t\tparentSignal?.aborted ||\n\t\t\t\ttimeoutSignal?.aborted ||\n\t\t\t\t(error instanceof Error &&\n\t\t\t\t\terror.name === \"AbortError\" &&\n\t\t\t\t\t!attemptTimedOut &&\n\t\t\t\t\ttimeoutSignal === undefined) ||\n\t\t\t\tattempt >= maxRetries\n\t\t\t) {\n\t\t\t\tthrow error;\n\t\t\t}\n\t\t}\n\t}\n}\n"]}

package/dist/utils/management-http.js

@@ -7,23 +7,23 @@ const RETRYABLE_STATUS_CODES = new Set([408, 425, 429, 500, 502, 503, 504]);
  * agent/model operations: those can fail after the HTTP request starts and are
  * retried by their semantic caller instead.
  *
- * Caller cancellation is terminal. When timeoutMs is supplied, it is the
- * overall time budget shared by all attempts.
+ * Caller cancellation and timeoutMs are terminal. attemptTimeoutMs aborts
+ * only the current attempt so a hung connection can be retried.
  */
 export async function fetchWithRetry(input, init = undefined, options = {}) {
     const maxRetries = options.maxRetries === undefined || !Number.isFinite(options.maxRetries)
         ? 2
         : Math.max(0, Math.floor(options.maxRetries));
     const retryOnStatus = options.retryOnStatus ?? true;
-    const parentSignal = init?.signal;
+    const parentSignal = init?.signal ?? undefined;
     const timeoutSignal = options.timeoutMs !== undefined && options.timeoutMs > 0 ? AbortSignal.timeout(options.timeoutMs) : undefined;
-    const signal = timeoutSignal
-        ? parentSignal
-            ? AbortSignal.any([parentSignal, timeoutSignal])
-            : timeoutSignal
-        : parentSignal;
+    const attemptTimeoutMs = options.attemptTimeoutMs !== undefined && options.attemptTimeoutMs > 0 ? options.attemptTimeoutMs : undefined;
     for (let attempt = 0;; attempt++) {
-        signal?.throwIfAborted();
+        parentSignal?.throwIfAborted();
+        timeoutSignal?.throwIfAborted();
+        const attemptTimeoutSignal = attemptTimeoutMs ? AbortSignal.timeout(attemptTimeoutMs) : undefined;
+        const signals = [parentSignal, timeoutSignal, attemptTimeoutSignal].filter((signal) => signal !== undefined);
+        const signal = signals.length > 1 ? AbortSignal.any(signals) : signals[0];
         try {
             const response = await fetch(input, signal ? { ...init, signal } : init);
             const shouldRetry = retryOnStatus && RETRYABLE_STATUS_CODES.has(response.status) && attempt < maxRetries;
@@ -38,9 +38,13 @@ export async function fetchWithRetry(input, init = undefined, options = {}) {
             }
         }
         catch (error) {
+            const attemptTimedOut = attemptTimeoutSignal?.aborted === true && !parentSignal?.aborted && !timeoutSignal?.aborted;
             if (parentSignal?.aborted ||
                 timeoutSignal?.aborted ||
-                (error instanceof Error && error.name === "AbortError" && timeoutSignal === undefined) ||
+                (error instanceof Error &&
+                    error.name === "AbortError" &&
+                    !attemptTimedOut &&
+                    timeoutSignal === undefined) ||
                 attempt >= maxRetries) {
                 throw error;
             }

package/dist/utils/management-http.js.map

@@ -1 +1 @@
-{"version":3,"file":"management-http.js","sourceRoot":"","sources":["../../src/utils/management-http.ts"],"names":[],"mappings":"AAEA,MAAM,sBAAsB,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAC,CAAC;AAW5E;;;;;;;;;;GAUG;AACH,MAAM,CAAC,KAAK,UAAU,cAAc,CACnC,KAAiB,EACjB,IAAI,GAA4B,SAAS,EACzC,OAAO,GAAsB,EAAE,EACX;IACpB,MAAM,UAAU,GACf,OAAO,CAAC,UAAU,KAAK,SAAS,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,UAAU,CAAC;QACvE,CAAC,CAAC,CAAC;QACH,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC,CAAC;IAChD,MAAM,aAAa,GAAG,OAAO,CAAC,aAAa,IAAI,IAAI,CAAC;IACpD,MAAM,YAAY,GAAG,IAAI,EAAE,MAAM,CAAC;IAClC,MAAM,aAAa,GAClB,OAAO,CAAC,SAAS,KAAK,SAAS,IAAI,OAAO,CAAC,SAAS,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,OAAO,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IAC/G,MAAM,MAAM,GAAG,aAAa;QAC3B,CAAC,CAAC,YAAY;YACb,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,YAAY,EAAE,aAAa,CAAC,CAAC;YAChD,CAAC,CAAC,aAAa;QAChB,CAAC,CAAC,YAAY,CAAC;IAEhB,KAAK,IAAI,OAAO,GAAG,CAAC,GAAI,OAAO,EAAE,EAAE,CAAC;QACnC,MAAM,EAAE,cAAc,EAAE,CAAC;QAEzB,IAAI,CAAC;YACJ,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC,EAAE,GAAG,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;YACzE,MAAM,WAAW,GAAG,aAAa,IAAI,sBAAsB,CAAC,GAAG,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,OAAO,GAAG,UAAU,CAAC;YACzG,IAAI,CAAC,WAAW;gBAAE,OAAO,QAAQ,CAAC;YAClC,IAAI,CAAC;gBACJ,MAAM,QAAQ,CAAC,IAAI,EAAE,MAAM,EAAE,CAAC;YAC/B,CAAC;YAAC,MAAM,CAAC;gBACR,6EAA6E;gBAC7E,wCAAwC;YACzC,CAAC;QACF,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YAChB,IACC,YAAY,EAAE,OAAO;gBACrB,aAAa,EAAE,OAAO;gBACtB,CAAC,KAAK,YAAY,KAAK,IAAI,KAAK,CAAC,IAAI,KAAK,YAAY,IAAI,aAAa,KAAK,SAAS,CAAC;gBACtF,OAAO,IAAI,UAAU,EACpB,CAAC;gBACF,MAAM,KAAK,CAAC;YACb,CAAC;QACF,CAAC;IACF,CAAC;AAAA,CACD","sourcesContent":["type FetchInput = Parameters<typeof fetch>[0];\n\nconst RETRYABLE_STATUS_CODES = new Set([408, 425, 429, 500, 502, 503, 504]);\n\nexport interface FetchRetryOptions {\n\t/** Number of additional attempts after the initial request. Defaults to two. */\n\tmaxRetries?: number;\n\t/** Retry transient HTTP responses as well as transport failures. Defaults to true. */\n\tretryOnStatus?: boolean;\n\t/** Per-attempt timeout. A new timeout is created for every attempt. */\n\ttimeoutMs?: number;\n}\n\n/**\n * Fetch a management HTTP resource with a bounded immediate retry.\n *\n * This is intentionally a transport-level helper for idempotent management\n * requests (version checks, catalogs, and downloads). It must not be used for\n * agent/model operations: those can fail after the HTTP request starts and are\n * retried by their semantic caller instead.\n *\n * Caller cancellation is terminal. When timeoutMs is supplied, it is the\n * overall time budget shared by all attempts.\n */\nexport async function fetchWithRetry(\n\tinput: FetchInput,\n\tinit: RequestInit | undefined = undefined,\n\toptions: FetchRetryOptions = {},\n): Promise<Response> {\n\tconst maxRetries =\n\t\toptions.maxRetries === undefined || !Number.isFinite(options.maxRetries)\n\t\t\t? 2\n\t\t\t: Math.max(0, Math.floor(options.maxRetries));\n\tconst retryOnStatus = options.retryOnStatus ?? true;\n\tconst parentSignal = init?.signal;\n\tconst timeoutSignal =\n\t\toptions.timeoutMs !== undefined && options.timeoutMs > 0 ? AbortSignal.timeout(options.timeoutMs) : undefined;\n\tconst signal = timeoutSignal\n\t\t? parentSignal\n\t\t\t? AbortSignal.any([parentSignal, timeoutSignal])\n\t\t\t: timeoutSignal\n\t\t: parentSignal;\n\n\tfor (let attempt = 0; ; attempt++) {\n\t\tsignal?.throwIfAborted();\n\n\t\ttry {\n\t\t\tconst response = await fetch(input, signal ? { ...init, signal } : init);\n\t\t\tconst shouldRetry = retryOnStatus && RETRYABLE_STATUS_CODES.has(response.status) && attempt < maxRetries;\n\t\t\tif (!shouldRetry) return response;\n\t\t\ttry {\n\t\t\t\tawait response.body?.cancel();\n\t\t\t} catch {\n\t\t\t\t// The response is being discarded before a retry. There is nothing useful to\n\t\t\t\t// do if cancelling its body also fails.\n\t\t\t}\n\t\t} catch (error) {\n\t\t\tif (\n\t\t\t\tparentSignal?.aborted ||\n\t\t\t\ttimeoutSignal?.aborted ||\n\t\t\t\t(error instanceof Error && error.name === \"AbortError\" && timeoutSignal === undefined) ||\n\t\t\t\tattempt >= maxRetries\n\t\t\t) {\n\t\t\t\tthrow error;\n\t\t\t}\n\t\t}\n\t}\n}\n"]}
+{"version":3,"file":"management-http.js","sourceRoot":"","sources":["../../src/utils/management-http.ts"],"names":[],"mappings":"AAEA,MAAM,sBAAsB,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAC,CAAC;AAa5E;;;;;;;;;;GAUG;AACH,MAAM,CAAC,KAAK,UAAU,cAAc,CACnC,KAAiB,EACjB,IAAI,GAA4B,SAAS,EACzC,OAAO,GAAsB,EAAE,EACX;IACpB,MAAM,UAAU,GACf,OAAO,CAAC,UAAU,KAAK,SAAS,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,UAAU,CAAC;QACvE,CAAC,CAAC,CAAC;QACH,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC,CAAC;IAChD,MAAM,aAAa,GAAG,OAAO,CAAC,aAAa,IAAI,IAAI,CAAC;IACpD,MAAM,YAAY,GAAG,IAAI,EAAE,MAAM,IAAI,SAAS,CAAC;IAC/C,MAAM,aAAa,GAClB,OAAO,CAAC,SAAS,KAAK,SAAS,IAAI,OAAO,CAAC,SAAS,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,OAAO,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IAC/G,MAAM,gBAAgB,GACrB,OAAO,CAAC,gBAAgB,KAAK,SAAS,IAAI,OAAO,CAAC,gBAAgB,GAAG,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,gBAAgB,CAAC,CAAC,CAAC,SAAS,CAAC;IAE/G,KAAK,IAAI,OAAO,GAAG,CAAC,GAAI,OAAO,EAAE,EAAE,CAAC;QACnC,YAAY,EAAE,cAAc,EAAE,CAAC;QAC/B,aAAa,EAAE,cAAc,EAAE,CAAC;QAChC,MAAM,oBAAoB,GAAG,gBAAgB,CAAC,CAAC,CAAC,WAAW,CAAC,OAAO,CAAC,gBAAgB,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QAClG,MAAM,OAAO,GAAG,CAAC,YAAY,EAAE,aAAa,EAAE,oBAAoB,CAAC,CAAC,MAAM,CACzE,CAAC,MAAM,EAAyB,EAAE,CAAC,MAAM,KAAK,SAAS,CACvD,CAAC;QACF,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;QAE1E,IAAI,CAAC;YACJ,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC,EAAE,GAAG,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;YACzE,MAAM,WAAW,GAAG,aAAa,IAAI,sBAAsB,CAAC,GAAG,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,OAAO,GAAG,UAAU,CAAC;YACzG,IAAI,CAAC,WAAW;gBAAE,OAAO,QAAQ,CAAC;YAClC,IAAI,CAAC;gBACJ,MAAM,QAAQ,CAAC,IAAI,EAAE,MAAM,EAAE,CAAC;YAC/B,CAAC;YAAC,MAAM,CAAC;gBACR,6EAA6E;gBAC7E,wCAAwC;YACzC,CAAC;QACF,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YAChB,MAAM,eAAe,GACpB,oBAAoB,EAAE,OAAO,KAAK,IAAI,IAAI,CAAC,YAAY,EAAE,OAAO,IAAI,CAAC,aAAa,EAAE,OAAO,CAAC;YAC7F,IACC,YAAY,EAAE,OAAO;gBACrB,aAAa,EAAE,OAAO;gBACtB,CAAC,KAAK,YAAY,KAAK;oBACtB,KAAK,CAAC,IAAI,KAAK,YAAY;oBAC3B,CAAC,eAAe;oBAChB,aAAa,KAAK,SAAS,CAAC;gBAC7B,OAAO,IAAI,UAAU,EACpB,CAAC;gBACF,MAAM,KAAK,CAAC;YACb,CAAC;QACF,CAAC;IACF,CAAC;AAAA,CACD","sourcesContent":["type FetchInput = Parameters<typeof fetch>[0];\n\nconst RETRYABLE_STATUS_CODES = new Set([408, 425, 429, 500, 502, 503, 504]);\n\nexport interface FetchRetryOptions {\n\t/** Number of additional attempts after the initial request. Defaults to two. */\n\tmaxRetries?: number;\n\t/** Retry transient HTTP responses as well as transport failures. Defaults to true. */\n\tretryOnStatus?: boolean;\n\t/** Overall time budget shared by all attempts. */\n\ttimeoutMs?: number;\n\t/** Per-attempt timeout. A new timeout is created for every attempt. */\n\tattemptTimeoutMs?: number;\n}\n\n/**\n * Fetch a management HTTP resource with a bounded immediate retry.\n *\n * This is intentionally a transport-level helper for idempotent management\n * requests (version checks, catalogs, and downloads). It must not be used for\n * agent/model operations: those can fail after the HTTP request starts and are\n * retried by their semantic caller instead.\n *\n * Caller cancellation and timeoutMs are terminal. attemptTimeoutMs aborts\n * only the current attempt so a hung connection can be retried.\n */\nexport async function fetchWithRetry(\n\tinput: FetchInput,\n\tinit: RequestInit | undefined = undefined,\n\toptions: FetchRetryOptions = {},\n): Promise<Response> {\n\tconst maxRetries =\n\t\toptions.maxRetries === undefined || !Number.isFinite(options.maxRetries)\n\t\t\t? 2\n\t\t\t: Math.max(0, Math.floor(options.maxRetries));\n\tconst retryOnStatus = options.retryOnStatus ?? true;\n\tconst parentSignal = init?.signal ?? undefined;\n\tconst timeoutSignal =\n\t\toptions.timeoutMs !== undefined && options.timeoutMs > 0 ? AbortSignal.timeout(options.timeoutMs) : undefined;\n\tconst attemptTimeoutMs =\n\t\toptions.attemptTimeoutMs !== undefined && options.attemptTimeoutMs > 0 ? options.attemptTimeoutMs : undefined;\n\n\tfor (let attempt = 0; ; attempt++) {\n\t\tparentSignal?.throwIfAborted();\n\t\ttimeoutSignal?.throwIfAborted();\n\t\tconst attemptTimeoutSignal = attemptTimeoutMs ? AbortSignal.timeout(attemptTimeoutMs) : undefined;\n\t\tconst signals = [parentSignal, timeoutSignal, attemptTimeoutSignal].filter(\n\t\t\t(signal): signal is AbortSignal => signal !== undefined,\n\t\t);\n\t\tconst signal = signals.length > 1 ? AbortSignal.any(signals) : signals[0];\n\n\t\ttry {\n\t\t\tconst response = await fetch(input, signal ? { ...init, signal } : init);\n\t\t\tconst shouldRetry = retryOnStatus && RETRYABLE_STATUS_CODES.has(response.status) && attempt < maxRetries;\n\t\t\tif (!shouldRetry) return response;\n\t\t\ttry {\n\t\t\t\tawait response.body?.cancel();\n\t\t\t} catch {\n\t\t\t\t// The response is being discarded before a retry. There is nothing useful to\n\t\t\t\t// do if cancelling its body also fails.\n\t\t\t}\n\t\t} catch (error) {\n\t\t\tconst attemptTimedOut =\n\t\t\t\tattemptTimeoutSignal?.aborted === true && !parentSignal?.aborted && !timeoutSignal?.aborted;\n\t\t\tif (\n\t\t\t\tparentSignal?.aborted ||\n\t\t\t\ttimeoutSignal?.aborted ||\n\t\t\t\t(error instanceof Error &&\n\t\t\t\t\terror.name === \"AbortError\" &&\n\t\t\t\t\t!attemptTimedOut &&\n\t\t\t\t\ttimeoutSignal === undefined) ||\n\t\t\t\tattempt >= maxRetries\n\t\t\t) {\n\t\t\t\tthrow error;\n\t\t\t}\n\t\t}\n\t}\n}\n"]}

package/dist/utils/text.d.ts

@@ -0,0 +1,8 @@
+/** Split a leading UTF-8 byte order mark from decoded text. */
+export declare function splitBom(content: string): {
+    bom: string;
+    text: string;
+};
+/** Remove a leading UTF-8 byte order mark from decoded text. */
+export declare function stripBom(content: string): string;
+//# sourceMappingURL=text.d.ts.map

package/dist/utils/text.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"text.d.ts","sourceRoot":"","sources":["../../src/utils/text.ts"],"names":[],"mappings":"AAAA,+DAA+D;AAC/D,wBAAgB,QAAQ,CAAC,OAAO,EAAE,MAAM,GAAG;IAAE,GAAG,EAAE,MAAM,CAAC;IAAC,IAAI,EAAE,MAAM,CAAA;CAAE,CAEvE;AAED,gEAAgE;AAChE,wBAAgB,QAAQ,CAAC,OAAO,EAAE,MAAM,GAAG,MAAM,CAEhD","sourcesContent":["/** Split a leading UTF-8 byte order mark from decoded text. */\nexport function splitBom(content: string): { bom: string; text: string } {\n\treturn content.startsWith(\"\\uFEFF\") ? { bom: \"\\uFEFF\", text: content.slice(1) } : { bom: \"\", text: content };\n}\n\n/** Remove a leading UTF-8 byte order mark from decoded text. */\nexport function stripBom(content: string): string {\n\treturn splitBom(content).text;\n}\n"]}

package/dist/utils/text.js

@@ -0,0 +1,9 @@
+/** Split a leading UTF-8 byte order mark from decoded text. */
+export function splitBom(content) {
+    return content.startsWith("\uFEFF") ? { bom: "\uFEFF", text: content.slice(1) } : { bom: "", text: content };
+}
+/** Remove a leading UTF-8 byte order mark from decoded text. */
+export function stripBom(content) {
+    return splitBom(content).text;
+}
+//# sourceMappingURL=text.js.map

package/dist/utils/text.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"text.js","sourceRoot":"","sources":["../../src/utils/text.ts"],"names":[],"mappings":"AAAA,+DAA+D;AAC/D,MAAM,UAAU,QAAQ,CAAC,OAAe,EAAiC;IACxE,OAAO,OAAO,CAAC,UAAU,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,QAAQ,EAAE,IAAI,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC;AAAA,CAC7G;AAED,gEAAgE;AAChE,MAAM,UAAU,QAAQ,CAAC,OAAe,EAAU;IACjD,OAAO,QAAQ,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC;AAAA,CAC9B","sourcesContent":["/** Split a leading UTF-8 byte order mark from decoded text. */\nexport function splitBom(content: string): { bom: string; text: string } {\n\treturn content.startsWith(\"\\uFEFF\") ? { bom: \"\\uFEFF\", text: content.slice(1) } : { bom: \"\", text: content };\n}\n\n/** Remove a leading UTF-8 byte order mark from decoded text. */\nexport function stripBom(content: string): string {\n\treturn splitBom(content).text;\n}\n"]}

package/docs/compaction.md

@@ -346,6 +346,21 @@ pi.on("session_before_compact", async (event, ctx) => {
 
 See [custom-compaction.ts](../examples/extensions/custom-compaction.ts) for a complete example using a different model.
 
+### session_compact_failed
+
+Fired when manual or automatic compaction fails or is aborted. This is useful for telemetry extensions that need to pair `session_before_compact` attempts with terminal outcomes.
+
+```typescript
+pi.on("session_compact_failed", async (event, ctx) => {
+  const { reason, errorMessage, aborted, willRetry, fromExtension } = event;
+  // reason - "manual" (/compact), "threshold", or "overflow"
+  // errorMessage - present for non-abort failures
+  // aborted - true for cancelled/aborted compactions
+  // willRetry - whether the aborted turn would have retried after compaction
+  // fromExtension - whether extension-provided compaction content was being used
+});
+```
+
 ### session_before_tree
 
 Fired before `/tree` navigation. Always fires regardless of whether user chose to summarize. Can cancel navigation or provide custom summary.

package/docs/custom-provider.md

@@ -752,8 +752,10 @@ interface ProviderModelConfig {
     requiresThinkingAsText?: boolean;
     requiresReasoningContentOnAssistantMessages?: boolean;
     thinkingFormat?: "openai" | "openrouter" | "deepseek" | "together" | "baseten" | "zai" | "qwen" | "chat-template" | "qwen-chat-template" | "string-thinking" | "ant-ling";
-    chatTemplateKwargs?: Record<string, string | number | boolean | null | { "$var": "thinking.enabled" | "thinking.effort"; omitWhenOff?: boolean }>;
-    chatTemplateArgs?: Record<string, string | number | boolean | null | { "$var": "thinking.enabled" | "thinking.effort"; omitWhenOff?: boolean }>;
+    chatTemplateKwargs?: Record<string, string | number | boolean | null | { "$var": "thinking.enabled" | "thinking.effort" | "thinking.budget"; omitWhenOff?: boolean }>;
+    chatTemplateArgs?: Record<string, string | number | boolean | null | { "$var": "thinking.enabled" | "thinking.effort" | "thinking.budget"; omitWhenOff?: boolean }>;
+    thinkingTokenBudgetField?: "thinking_token_budget" | "thinking_budget" | "thinking_budget_tokens";
+    supportsThinkingTokenBudget?: boolean;
     cacheControlFormat?: "anthropic";
     sessionAffinityFormat?: "openai" | "openai-nosession" | "openrouter";
     sendSessionAffinityHeaders?: boolean;
@@ -771,4 +773,5 @@ interface ProviderModelConfig {
 ```
 
 `openrouter` sends `reasoning: { effort }`. `deepseek` sends `thinking: { type: "enabled" | "disabled" }` and `reasoning_effort` when enabled. `together` sends `reasoning: { enabled }` and also `reasoning_effort` when `supportsReasoningEffort` is enabled. `qwen` is for DashScope-style top-level `enable_thinking`. Use `qwen-chat-template` for local Qwen-compatible servers that read `chat_template_kwargs.enable_thinking` and need `preserve_thinking`. Use `chat-template` for configurable `chat_template_kwargs`, for example DeepSeek V3.x behind vLLM with `chatTemplateKwargs: { "thinking": { "$var": "thinking.enabled" } }`. Use `thinkingFormat: "baseten"` with `chatTemplateArgs` when the provider expects toggle values under `chat_template_args` and optionally supports top-level `reasoning_effort`.
+`thinkingTokenBudgetField` sends a clamped per-level thinking budget as a top-level request field (`thinking_token_budget` on vLLM, `thinking_budget` on Qwen/SGLang, `thinking_budget_tokens` on llama.cpp). `supportsThinkingTokenBudget: true` is an alias for the vLLM field name. Do not combine it with `reasoning_effort` on DashScope Qwen models.
 `cacheControlFormat: "anthropic"` applies Anthropic-style `cache_control` markers to the system prompt, last tool definition, and last user, assistant, or tool-result text content.

package/docs/extensions.md

@@ -330,7 +330,8 @@ user sends another prompt ◄─────────────────
 
 /compact or auto-compaction
   ├─► session_before_compact (can cancel or customize)
-  └─► session_compact
+  ├─► session_compact (success)
+  └─► session_compact_failed (failure or abort)
 
 /tree navigation
   ├─► session_before_tree (can cancel or customize)
@@ -448,7 +449,7 @@ pi.on("session_before_fork", async (event, ctx) => {
 After a successful fork or clone, pi emits `session_shutdown` for the old extension instance, reloads and rebinds extensions for the new session, then emits `session_start` with `reason: "fork"` and `previousSessionFile`.
 Do cleanup work in `session_shutdown`, then reestablish any in-memory state in `session_start`.
 
-#### session_before_compact / session_compact
+#### session_before_compact / session_compact / session_compact_failed
 
 Fired on compaction. See [compaction.md](compaction.md) for details.
 
@@ -479,6 +480,14 @@ pi.on("session_compact", async (event, ctx) => {
   // event.reason - "manual" (/compact), "threshold", or "overflow"
   // event.willRetry - whether the aborted turn is retried after compaction (overflow recovery)
 });
+
+pi.on("session_compact_failed", async (event, ctx) => {
+  // event.reason - "manual" (/compact), "threshold", or "overflow"
+  // event.errorMessage - present for non-abort failures
+  // event.aborted - true for cancelled/aborted compactions
+  // event.willRetry - whether the aborted turn would have retried after compaction
+  // event.fromExtension - whether extension-provided compaction content was being used
+});
 ```
 
 #### session_before_tree / session_tree

package/docs/json.md

@@ -15,12 +15,16 @@ except that streaming message updates omit cumulative snapshots:
 ```typescript
 type WithoutPartial<T> = T extends { partial: unknown } ? Omit<T, "partial"> : T;
 
+type JsonAssistantMessageEvent<T> = T extends { type: "toolcall_start"; partial: unknown }
+  ? WithoutPartial<T> & { id: string; toolName: string }
+  : WithoutPartial<T>;
+
 type JsonAgentSessionEvent =
   | Exclude<AgentSessionEvent, { type: "message_update" }>
   | {
       type: "message_update";
       usage: Usage;
-      assistantMessageEvent: WithoutPartial<AssistantMessageEvent>;
+      assistantMessageEvent: JsonAssistantMessageEvent<AssistantMessageEvent>;
     };
 ```
 
@@ -84,7 +88,8 @@ Followed by events as they occur:
 `assistantMessageEvent.partial` to keep stream size linear. The top-level `usage` field contains
 the latest cumulative provider-reported usage and may remain zero when a provider only reports
 usage at completion. Use `contentIndex` and `delta` to assemble live text, thinking, or tool-call
-arguments if needed. `message_end` contains the final authoritative message.
+arguments if needed. A `toolcall_start` event also includes the constant-sized `id` and `toolName`
+fields. `message_end` contains the final authoritative message.
 
 ## Example
 

package/docs/llama-cpp.md

@@ -53,6 +53,8 @@ Start Pi and configure the provider:
 
 Enter the router URL and optional API key. The default URL is `http://127.0.0.1:8080`.
 
+If you start the router with `--no-models-autoload`, `/login llama.cpp` only stores the connection. Run `/llama` to load a model, then `/model` to select the loaded model for the current session.
+
 Environment variables can configure the same values without `/login`:
 
 ```bash

package/docs/models.md

@@ -254,6 +254,8 @@ Current behavior:
 
 Only OpenAI-compatible APIs apply it (`openai-completions`, `openai-responses`, `azure-openai-responses`); other APIs ignore it. Keys override pi's named request fields (for example a `temperature` key here beats the request-level temperature), so prefer it as the single source of sampling truth for a model. In `modelOverrides`, `samplingParams` merges per key with the base model's value.
 
+A constant thinking-token cap can go here too, but it will not follow `thinkingBudgets` or leave room for the answer. Prefer `compat.thinkingTokenBudgetField` (or the `supportsThinkingTokenBudget` alias) for that.
+
 ### Thinking Level Map
 
 Use `thinkingLevelMap` on a model to describe model-specific thinking controls. Keys are pi thinking levels: `off`, `minimal`, `low`, `medium`, `high`, `xhigh`, `max`. Maps may contain holes; for example, a model can expose `high` and `max` without exposing `xhigh`.
@@ -467,8 +469,10 @@ For providers with partial OpenAI compatibility, use the `compat` field.
 | `requiresThinkingAsText` | Convert thinking blocks to plain text |
 | `requiresReasoningContentOnAssistantMessages` | Include empty `reasoning_content` on all replayed assistant messages when reasoning is enabled |
 | `thinkingFormat` | Use `reasoning_effort`, `openrouter`, `deepseek`, `together`, `baseten`, `zai`, `qwen`, `chat-template`, or `qwen-chat-template` thinking parameters |
-| `chatTemplateKwargs` | `chat_template_kwargs` values for `thinkingFormat: "chat-template"`; use `{ "$var": "thinking.enabled" }` or `{ "$var": "thinking.effort" }` for pi-controlled thinking values |
-| `chatTemplateArgs` | `chat_template_args` values for `thinkingFormat: "baseten"`; use `{ "$var": "thinking.enabled" }` or `{ "$var": "thinking.effort" }` for pi-controlled thinking values |
+| `chatTemplateKwargs` | `chat_template_kwargs` values for `thinkingFormat: "chat-template"`; use `{ "$var": "thinking.enabled" }`, `{ "$var": "thinking.effort" }`, or `{ "$var": "thinking.budget" }` for pi-controlled thinking values |
+| `chatTemplateArgs` | `chat_template_args` values for `thinkingFormat: "baseten"`; use `{ "$var": "thinking.enabled" }`, `{ "$var": "thinking.effort" }`, or `{ "$var": "thinking.budget" }` for pi-controlled thinking values |
+| `thinkingTokenBudgetField` | Top-level request field used to cap reasoning tokens from `thinkingBudgets`, clamped so at least 1024 tokens remain for the answer. `"thinking_token_budget"` (vLLM), `"thinking_budget"` (Qwen/DashScope/SGLang), `"thinking_budget_tokens"` (llama.cpp). Off by default; not set on the generated catalog. |
+| `supportsThinkingTokenBudget` | Alias for `thinkingTokenBudgetField: "thinking_token_budget"` (vLLM). Prefer `thinkingTokenBudgetField`. Default: `false`. |
 | `cacheControlFormat` | Use Anthropic-style `cache_control` markers on the system prompt, last tool definition, and last user, assistant, or tool-result text content. Currently only `anthropic` is supported. |
 | `sendSessionAffinityHeaders` | For `openai-completions`, send session-affinity headers from the session id when caching is enabled. Default: `false`. |
 | `sessionAffinityFormat` | For `openai-completions` and `openai-responses`, the session-affinity header format: `openai` sends `session_id`/`x-client-request-id` (completions also `x-session-affinity`), `openai-nosession` omits the underscore-containing `session_id` header, `openrouter` sends `x-session-id`. Does not affect the `prompt_cache_key` body param. Default: auto-detected. |
@@ -481,6 +485,8 @@ For providers with partial OpenAI compatibility, use the `compat` field.
 
 `openrouter` uses `reasoning: { effort }`. `together` uses `reasoning: { enabled }` and also `reasoning_effort` when `supportsReasoningEffort` is enabled. `qwen` uses top-level `enable_thinking`. Use `qwen-chat-template` for local Qwen-compatible servers that require `chat_template_kwargs.enable_thinking` and `preserve_thinking`. Use `chat-template` for vLLM/Hugging Face chat templates that need configurable `chat_template_kwargs`, such as `chatTemplateKwargs: { "thinking": { "$var": "thinking.enabled" } }` for DeepSeek V3.x templates. Use `thinkingFormat: "baseten"` with `chatTemplateArgs` for providers that expose toggle controls through `chat_template_args` and optionally support top-level `reasoning_effort`.
 
+`thinkingTokenBudgetField` is independent of `thinkingFormat`. Do not enable it on the generated Qwen catalog: those models already send `reasoning_effort`, and DashScope rejects `thinking_budget` together with `reasoning_effort`.
+
 `cacheControlFormat: "anthropic"` is for OpenAI-compatible providers that expose Anthropic-style prompt caching through `cache_control` markers on text content and tool definitions.
 
 Example:

package/docs/rpc.md

@@ -945,7 +945,7 @@ The `assistantMessageEvent` field contains one of these delta types:
 | `thinking_start` | Thinking block started |
 | `thinking_delta` | Thinking content chunk |
 | `thinking_end` | Thinking block ended |
-| `toolcall_start` | Tool call started |
+| `toolcall_start` | Tool call started (includes `id` and `toolName`) |
 | `toolcall_delta` | Tool call arguments chunk |
 | `toolcall_end` | Tool call ended (includes full `toolCall` object) |
 
@@ -960,11 +960,17 @@ Example streaming a text response:
 The top-level `usage` field contains the latest cumulative provider-reported usage. It may remain
 zero until completion when a provider does not report usage during streaming.
 
+Example starting a tool call:
+```json
+{"type":"message_update","usage":{...},"assistantMessageEvent":{"type":"toolcall_start","contentIndex":1,"id":"call_abc123","toolName":"write"}}
+```
+
 `message_update` intentionally omits the former cumulative `message` field and
 `assistantMessageEvent.partial`. Clients that need a live partial message must assemble it
 from `message_start` and subsequent events using `contentIndex`. Treat `message_end.message`
-as authoritative. For tool calls, buffer `toolcall_delta.delta`; `toolcall_end.toolCall`
-contains the completed call.
+as authoritative. For tool calls, `toolcall_start` provides the call `id` and `toolName`;
+buffer `toolcall_delta.delta` for arguments. `toolcall_end.toolCall` contains the completed
+call.
 
 ### bash_execution_update
 

package/docs/settings.md

@@ -31,8 +31,8 @@ Use `/trust` in interactive mode to save a project trust decision for future ses
 | `defaultModel` | string | - | Default model ID |
 | `defaultThinkingLevel` | string | - | `"off"`, `"minimal"`, `"low"`, `"medium"`, `"high"`, `"xhigh"`, `"max"` |
 | `hideThinkingBlock` | boolean | `false` | Hide thinking blocks in output |
-| `showCacheMissNotices` | boolean | `false` | Show transcript notices for significant prompt-cache misses |
-| `thinkingBudgets` | object | - | Custom token budgets per thinking level |
+| `showCacheMissNotices` | boolean | `false` | Show transcript notices for significant prompt-cache misses and compaction or branch-summary usage |
+| `thinkingBudgets` | object | - | Custom token budgets per thinking level. Anthropic, Google, and Bedrock use these natively. OpenAI-compatible models use them when `compat.thinkingTokenBudgetField` (or `supportsThinkingTokenBudget`) is set. |
 
 #### thinkingBudgets
 

package/docs/skills.md

@@ -34,9 +34,10 @@ Pi loads skills from:
 - CLI: `--skill <path>` (repeatable, additive even with `--no-skills`)
 
 Discovery rules:
-- In `~/.pi/agent/skills/` and `.pi/skills/`, direct root `.md` files are discovered as individual skills
+- In `~/.pi/agent/skills/` and `.pi/skills/`, direct root `.md` files are discovered as individual skills when they have valid skill frontmatter with a non-empty `description`
 - In all skill locations, directories containing `SKILL.md` are discovered recursively
-- In `~/.agents/skills/` and project `.agents/skills/`, root `.md` files are ignored
+- In `~/.agents/skills/` and project `.agents/skills/`, root `.md` files are ignored, but nested `.md` files in grouping folders are discovered when they declare skill frontmatter
+- Root Markdown files other than `SKILL.md` that do not look like skills are ignored silently
 
 Disable discovery with `--no-skills` (explicit `--skill` paths still load).
 
@@ -183,7 +184,7 @@ Pi validates skills against the Agent Skills standard. Most issues produce warni
 
 Unknown frontmatter fields are ignored.
 
-**Exception:** Skills with missing description are not loaded.
+Declared skills with missing descriptions are not loaded. Malformed `SKILL.md` files and `SKILL.md` files without a description produce warnings and are not loaded. Other Markdown files without valid skill frontmatter are ignored.
 
 Name collisions (same name from different locations) warn and keep the first skill found.
 

package/examples/extensions/border-status-editor.ts

@@ -92,7 +92,7 @@ export default function (pi: ExtensionAPI) {
 		activeTui?.requestRender();
 	});
 
-	pi.on("agent_end", () => {
+	pi.on("agent_settled", () => {
 		isWorking = false;
 		stopSpinner();
 		activeTui?.requestRender();

package/examples/extensions/git-checkpoint.ts

@@ -46,8 +46,8 @@ export default function (pi: ExtensionAPI) {
 		}
 	});
 
-	pi.on("agent_end", async () => {
-		// Clear checkpoints after agent completes
+	pi.on("agent_settled", async () => {
+		// Clear checkpoints after the full agent run completes
 		checkpoints.clear();
 	});
 }

package/examples/extensions/notify.ts

@@ -49,7 +49,9 @@ function notify(title: string, body: string): void {
 }
 
 export default function (pi: ExtensionAPI) {
-	pi.on("agent_end", async () => {
+	// `agent_end` fires after each low-level run; Pi may still retry, compact,
+	// or continue with queued follow-ups. Notify only after the full run settles.
+	pi.on("agent_settled", async () => {
 		notify("Pi", "Ready for input");
 	});
 }

package/examples/extensions/subagent/README.md

@@ -62,7 +62,7 @@ This tool executes a separate `pi` subprocess with a delegated system prompt and
 
 To enable project-local agents, pass `agentScope: "both"` (or `"project"`). Only do this for repositories you trust.
 
-When running interactively, the tool prompts for confirmation before running project-local agents. Set `confirmProjectAgents: false` to disable.
+When running interactively, the tool prompts for confirmation before running project-local agents in untrusted projects. Trusted projects skip the additional prompt. Set `confirmProjectAgents: false` to disable confirmation.
 
 ## Usage
 

package/examples/extensions/subagent/index.ts

@@ -517,7 +517,12 @@ export default function (pi: ExtensionAPI) {
 				};
 			}
 
-			if ((agentScope === "project" || agentScope === "both") && confirmProjectAgents && ctx.hasUI) {
+			if (
+				(agentScope === "project" || agentScope === "both") &&
+				confirmProjectAgents &&
+				ctx.hasUI &&
+				!ctx.isProjectTrusted()
+			) {
 				const requestedAgentNames = new Set<string>();
 				if (params.chain) for (const step of params.chain) requestedAgentNames.add(step.agent);
 				if (params.tasks) for (const t of params.tasks) requestedAgentNames.add(t.agent);

package/examples/extensions/titlebar-spinner.ts

@@ -48,7 +48,7 @@ export default function (pi: ExtensionAPI) {
 		startAnimation(ctx);
 	});
 
-	pi.on("agent_end", async (_event, ctx) => {
+	pi.on("agent_settled", async (_event, ctx) => {
 		stopAnimation(ctx);
 	});
 

package/examples/rpc-extension-ui.ts

@@ -576,7 +576,7 @@ async function main() {
 			return;
 		}
 
-		if (data.type === "agent_end") {
+		if (data.type === "agent_settled") {
 			isStreaming = false;
 			hideLoading();
 			outputLog.append("");

package/examples/sdk/06-extensions.ts

@@ -71,7 +71,7 @@ export default function (pi: ExtensionAPI) {
 	});
 
 	pi.on("agent_end", async (event) => {
-		console.log(\`[Extension] Done, \${event.messages.length} messages\`);
+		console.log(\`[Extension] Low-level run ended, \${event.messages.length} messages\`);
 	});
 
 	// Register a custom tool

package/examples/sdk/README.md

@@ -132,7 +132,7 @@ session.subscribe((event) => {
     case "tool_execution_end":
       console.log(`Result: ${event.result}`);
       break;
-    case "agent_end":
+    case "agent_settled":
       console.log("Done");
       break;
   }

package/npm-shrinkwrap.json

@@ -1,12 +1,12 @@
 {
 	"name": "@hoilab/ada-cli",
-	"version": "0.84.21",
+	"version": "0.84.22",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "@hoilab/ada-cli",
-			"version": "0.84.21",
+			"version": "0.84.22",
 			"license": "MIT",
 			"dependencies": {
 				"@earendil-works/pi-agent-core": "^0.84.2",
@@ -501,7 +501,6 @@
 				"@aws-sdk/client-bedrock-runtime": "3.1048.0",
 				"@earendil-works/pi-telemetry": "^0.84.2",
 				"@google/genai": "1.52.0",
-				"@opentelemetry/api": "1.9.0",
 				"@smithy/node-http-handler": "4.7.3",
 				"http-proxy-agent": "7.0.2",
 				"https-proxy-agent": "7.0.6",
@@ -773,15 +772,6 @@
 				}
 			]
 		},
-		"node_modules/@opentelemetry/api": {
-			"version": "1.9.0",
-			"resolved": "https://registry.npmjs.org/@opentelemetry/api/-/api-1.9.0.tgz",
-			"integrity": "sha512-3giAOQvZiH5F9bMlMiv8+GSPMeqg0dbaeo58/0SlA9sxSqZhnUtxzX9/2FzyhS9sWQf5S0GJE0AKBrFqjpeYcg==",
-			"license": "Apache-2.0",
-			"engines": {
-				"node": ">=8.0.0"
-			}
-		},
 		"node_modules/@protobufjs/aspromise": {
 			"version": "1.1.2",
 			"resolved": "https://registry.npmjs.org/@protobufjs/aspromise/-/aspromise-1.1.2.tgz",

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@hoilab/ada-cli",
-	"version": "0.84.21",
+	"version": "0.84.22",
 	"description": "Ada CLI - coding agent CLI (rebranded fork of pi.dev) with read, bash, edit, write tools and session management",
 	"type": "module",
 	"piConfig": {
@@ -37,7 +37,7 @@
 		"clean": "shx rm -rf dist",
 		"build": "tsgo -p tsconfig.build.json && shx chmod +x dist/cli.js dist/rpc-entry.js && npm run copy-assets",
 		"build:binary": "npm --prefix ../tui run build && npm --prefix ../telemetry run build && npm --prefix ../ai run build && npm --prefix ../agent run build && npm --prefix ../protocol run build && npm --prefix ../client run build && npm run build && bun build --compile --no-compile-autoload-bunfig ./dist/bun/cli.js ./src/utils/image-resize-worker.ts --outfile dist/ada-cli && npm run copy-binary-assets",
-		"copy-assets": "shx mkdir -p dist/modes/interactive/theme && shx cp src/modes/interactive/theme/*.json dist/modes/interactive/theme/ && shx mkdir -p dist/modes/interactive/assets && shx cp src/modes/interactive/assets/*.png dist/modes/interactive/assets/ && shx mkdir -p dist/core/export-html/vendor && shx cp src/core/export-html/template.html src/core/export-html/template.css src/core/export-html/template.js dist/core/export-html/ && shx cp src/core/export-html/vendor/*.js dist/core/export-html/vendor/",
+		"copy-assets": "shx mkdir -p dist/modes/interactive/theme && shx cp src/modes/interactive/theme/*.json dist/modes/interactive/theme/ && shx mkdir -p dist/modes/interactive/assets && shx cp src/modes/interactive/assets/*.png dist/modes/interactive/assets/ && shx mkdir -p dist/core/export-html/vendor && shx cp src/core/export-html/template.html src/core/export-html/template.css src/core/export-html/template.js dist/core/export-html/ && shx cp src/core/export-html/vendor/*.js dist/core/export-html/vendor/ && shx mkdir -p dist/skills/ada-engine && shx cp skills/ada-engine/SKILL.md dist/skills/ada-engine/",
 		"copy-binary-assets": "shx cp package.json dist/ && shx cp README.md dist/ && shx cp CHANGELOG.md dist/ && shx mkdir -p dist/theme && shx cp src/modes/interactive/theme/*.json dist/theme/ && shx mkdir -p dist/assets && shx cp src/modes/interactive/assets/*.png dist/assets/ && shx mkdir -p dist/export-html/vendor && shx cp src/core/export-html/template.html dist/export-html/ && shx cp src/core/export-html/vendor/*.js dist/export-html/vendor/ && shx cp -r docs dist/ && shx cp -r examples dist/ && shx cp ../../node_modules/@silvia-odwyer/photon-node/photon_rs_bg.wasm dist/",
 		"test": "vitest --run",
 		"shrinkwrap": "node ../../scripts/generate-coding-agent-shrinkwrap.mjs",
@@ -106,4 +106,4 @@
 	"engines": {
 		"node": ">=22.19.0"
 	}
-}
+}