@oneharness/sdk 0.5.9 → 0.5.10

package/dist/generated/history.d.ts

@@ -69,6 +69,7 @@ export type HistoryRecord = {
      * every record so a reader can resolve a session by name from any line.
      */
     name: string;
+    observed_tool_ms?: never | undefined;
     /**
      * The normalized approval mode requested for the run.
      */
@@ -83,7 +84,7 @@ export type HistoryRecord = {
      * run's single prompt).
      */
     prompt: string;
-    schema_version: "1.1";
+    schema_version: "1.2";
     /**
      * The oneharness session id this run belongs to (the history file's stem).
      */
@@ -117,7 +118,7 @@ export type HistoryRecord = {
      * Best-effort normalized tool-call events; `null` when the harness exposes
      * no machine-readable trace.
      */
-    events: ((ActionEvent & {
+    events: (((ActionEvent & {
         duration_ms: number;
         finished_at: string;
         kind: "tool_call";
@@ -148,7 +149,10 @@ export type HistoryRecord = {
     }) | (ActionEvent & {
         kind?: "tool_result" | undefined;
         [k: string]: unknown;
-    }))[] | null;
+    })) & {
+        timing_source?: never | undefined;
+        [k: string]: unknown;
+    })[] | null;
     exit_code: number | null;
     /**
      * Best-effort classified failure reason (see [`FailureKind`]); `null` when
@@ -177,6 +181,7 @@ export type HistoryRecord = {
      * every record so a reader can resolve a session by name from any line.
      */
     name: string;
+    observed_tool_ms?: never | undefined;
     /**
      * The normalized approval mode requested for the run.
      */
@@ -191,7 +196,7 @@ export type HistoryRecord = {
      * run's single prompt).
      */
     prompt: string;
-    schema_version: "1.0";
+    schema_version: "1.1";
     /**
      * The oneharness session id this run belongs to (the history file's stem).
      */
@@ -219,6 +224,118 @@ export type HistoryRecord = {
     usage: Usage1;
     variant?: string | null | undefined;
     [k: string]: unknown;
+} | {
+    duration_ms: number;
+    /**
+     * Best-effort normalized tool-call events; `null` when the harness exposes
+     * no machine-readable trace.
+     */
+    events: (((ActionEvent & {
+        duration_ms: number;
+        finished_at: string;
+        kind: "tool_call";
+        started_at: string;
+        status: "completed";
+        tool_call_id: string;
+        [k: string]: unknown;
+    }) | (ActionEvent & {
+        duration_ms: number;
+        finished_at: string;
+        kind: "tool_call";
+        started_at: string;
+        status: "failed";
+        tool_call_id: string;
+        [k: string]: unknown;
+    }) | (ActionEvent & {
+        kind: "tool_call";
+        started_at: string;
+        status: "timeout";
+        tool_call_id: string;
+        [k: string]: unknown;
+    }) | (ActionEvent & {
+        kind: "tool_call";
+        started_at: string;
+        status: "interrupted";
+        tool_call_id: string;
+        [k: string]: unknown;
+    }) | (ActionEvent & {
+        kind?: "tool_result" | undefined;
+        [k: string]: unknown;
+    })) & {
+        timing_source?: never | undefined;
+        [k: string]: unknown;
+    })[] | null;
+    exit_code: number | null;
+    /**
+     * Best-effort classified failure reason (see [`FailureKind`]); `null` when
+     * unclassified.
+     */
+    failure_kind: FailureKind | null;
+    finished_at: string | null;
+    /**
+     * Canonical harness id (e.g. `claude-code`).
+     */
+    harness: string;
+    harness_id?: string | undefined;
+    /**
+     * Globally unique, time-ordered record id. This is also the cursor accepted
+     * by `history watch --after` and the exact id accepted by history lookup.
+     */
+    history_id: string;
+    labels?: HistoryLabels2 | undefined;
+    /**
+     * The effective top-level model for the run, if any.
+     */
+    model: string | null;
+    model_ms: number;
+    /**
+     * The human-meaningful session name (see [`session_name`]); repeated on
+     * every record so a reader can resolve a session by name from any line.
+     */
+    name: string;
+    observed_tool_ms?: never | undefined;
+    /**
+     * The normalized approval mode requested for the run.
+     */
+    permission_mode: "read-only" | "plan" | "default" | "edit" | "auto" | "bypass";
+    /**
+     * The project directory the run operated in (the real path, not the
+     * on-disk slug), so the list view can show where a session ran.
+     */
+    project: string;
+    /**
+     * The prompt this harness run received (its own, on a batch run; else the
+     * run's single prompt).
+     */
+    prompt: string;
+    schema_version: "1.0";
+    /**
+     * The oneharness session id this run belongs to (the history file's stem).
+     */
+    session: string;
+    /**
+     * The harness's own continuation id, when it exposed one; `null` otherwise.
+     */
+    session_id: string | null;
+    started_at: string;
+    status: Status;
+    /**
+     * Best-effort final assistant text; `null` when extraction was impossible.
+     */
+    text: string | null;
+    /**
+     * How `text` was extracted; `null` when absent.
+     */
+    text_source: string | null;
+    time_to_first_token_ms?: number | null | undefined;
+    /**
+     * RFC3339 UTC instant the record was written (append time).
+     */
+    timestamp: string;
+    tool_ms: number;
+    usage: Usage2;
+    variant?: string | null | undefined;
+    [k: string]: unknown;
 } | {
     duration_ms: number | null;
     /**
@@ -255,7 +372,7 @@ export type HistoryRecord = {
      * by `history watch --after` and the exact id accepted by history lookup.
      */
     history_id: string;
-    labels?: HistoryLabels2 | undefined;
+    labels?: HistoryLabels3 | undefined;
     /**
      * The effective top-level model for the run, if any.
      */
@@ -266,6 +383,178 @@ export type HistoryRecord = {
      * every record so a reader can resolve a session by name from any line.
      */
     name: string;
+    observed_tool_ms?: never | undefined;
+    /**
+     * The normalized approval mode requested for the run.
+     */
+    permission_mode: "read-only" | "plan" | "default" | "edit" | "auto" | "bypass";
+    /**
+     * The project directory the run operated in (the real path, not the
+     * on-disk slug), so the list view can show where a session ran.
+     */
+    project: string;
+    /**
+     * The prompt this harness run received (its own, on a batch run; else the
+     * run's single prompt).
+     */
+    prompt: string;
+    schema_version: "1.2";
+    /**
+     * The oneharness session id this run belongs to (the history file's stem).
+     */
+    session: string;
+    /**
+     * The harness's own continuation id, when it exposed one; `null` otherwise.
+     */
+    session_id: string | null;
+    started_at?: never | undefined;
+    status: Status;
+    /**
+     * Best-effort final assistant text; `null` when extraction was impossible.
+     */
+    text: string | null;
+    /**
+     * How `text` was extracted; `null` when absent.
+     */
+    text_source: string | null;
+    time_to_first_token_ms?: never | undefined;
+    /**
+     * RFC3339 UTC instant the record was written (append time).
+     */
+    timestamp: string;
+    tool_ms?: never | undefined;
+    usage: Usage3;
+    variant?: string | null | undefined;
+    [k: string]: unknown;
+} | {
+    duration_ms: number;
+    /**
+     * Best-effort normalized tool-call events; `null` when the harness exposes
+     * no machine-readable trace.
+     */
+    events: ActionEvent[] | null;
+    exit_code: number | null;
+    /**
+     * Best-effort classified failure reason (see [`FailureKind`]); `null` when
+     * unclassified.
+     */
+    failure_kind: FailureKind | null;
+    finished_at: null;
+    /**
+     * Canonical harness id (e.g. `claude-code`).
+     */
+    harness: string;
+    harness_id: string;
+    /**
+     * Globally unique, time-ordered record id. This is also the cursor accepted
+     * by `history watch --after` and the exact id accepted by history lookup.
+     */
+    history_id: string;
+    labels?: HistoryLabels4 | undefined;
+    /**
+     * The effective top-level model for the run, if any.
+     */
+    model: string | null;
+    model_ms?: never | undefined;
+    /**
+     * The human-meaningful session name (see [`session_name`]); repeated on
+     * every record so a reader can resolve a session by name from any line.
+     */
+    name: string;
+    observed_tool_ms: number;
+    /**
+     * The normalized approval mode requested for the run.
+     */
+    permission_mode: "read-only" | "plan" | "default" | "edit" | "auto" | "bypass";
+    /**
+     * The project directory the run operated in (the real path, not the
+     * on-disk slug), so the list view can show where a session ran.
+     */
+    project: string;
+    /**
+     * The prompt this harness run received (its own, on a batch run; else the
+     * run's single prompt).
+     */
+    prompt: string;
+    schema_version: "1.2";
+    /**
+     * The oneharness session id this run belongs to (the history file's stem).
+     */
+    session: string;
+    /**
+     * The harness's own continuation id, when it exposed one; `null` otherwise.
+     */
+    session_id: string | null;
+    started_at?: never | undefined;
+    status: Status;
+    /**
+     * Best-effort final assistant text; `null` when extraction was impossible.
+     */
+    text: string | null;
+    /**
+     * How `text` was extracted; `null` when absent.
+     */
+    text_source: string | null;
+    time_to_first_token_ms?: never | undefined;
+    /**
+     * RFC3339 UTC instant the record was written (append time).
+     */
+    timestamp: string;
+    tool_ms?: never | undefined;
+    usage: Usage4;
+    variant?: string | null | undefined;
+    [k: string]: unknown;
+} | {
+    duration_ms: number | null;
+    /**
+     * Best-effort normalized tool-call events; `null` when the harness exposes
+     * no machine-readable trace.
+     */
+    events: ({
+        duration_ms?: null | undefined;
+        finished_at?: null | undefined;
+        index: number;
+        input: unknown;
+        kind: string;
+        name: string | null;
+        output: string | null;
+        started_at?: null | undefined;
+        status?: null | undefined;
+        tool_call_id?: string | null | undefined;
+        [k: string]: unknown;
+    } & {
+        timing_source?: never | undefined;
+        [k: string]: unknown;
+    })[] | null;
+    exit_code: number | null;
+    /**
+     * Best-effort classified failure reason (see [`FailureKind`]); `null` when
+     * unclassified.
+     */
+    failure_kind: FailureKind | null;
+    finished_at: null;
+    /**
+     * Canonical harness id (e.g. `claude-code`).
+     */
+    harness: string;
+    harness_id?: string | undefined;
+    /**
+     * Globally unique, time-ordered record id. This is also the cursor accepted
+     * by `history watch --after` and the exact id accepted by history lookup.
+     */
+    history_id: string;
+    labels?: HistoryLabels5 | undefined;
+    /**
+     * The effective top-level model for the run, if any.
+     */
+    model: string | null;
+    model_ms?: never | undefined;
+    /**
+     * The human-meaningful session name (see [`session_name`]); repeated on
+     * every record so a reader can resolve a session by name from any line.
+     */
+    name: string;
+    observed_tool_ms?: never | undefined;
     /**
      * The normalized approval mode requested for the run.
      */
@@ -305,7 +594,7 @@ export type HistoryRecord = {
      */
     timestamp: string;
     tool_ms?: never | undefined;
-    usage: Usage2;
+    usage: Usage5;
     variant?: string | null | undefined;
     [k: string]: unknown;
 } | {
@@ -314,7 +603,7 @@ export type HistoryRecord = {
      * Best-effort normalized tool-call events; `null` when the harness exposes
      * no machine-readable trace.
      */
-    events: {
+    events: ({
         duration_ms?: null | undefined;
         finished_at?: null | undefined;
         index: number;
@@ -326,7 +615,10 @@ export type HistoryRecord = {
         status?: null | undefined;
         tool_call_id?: string | null | undefined;
         [k: string]: unknown;
-    }[] | null;
+    } & {
+        timing_source?: never | undefined;
+        [k: string]: unknown;
+    })[] | null;
     exit_code: number | null;
     /**
      * Best-effort classified failure reason (see [`FailureKind`]); `null` when
@@ -344,7 +636,7 @@ export type HistoryRecord = {
      * by `history watch --after` and the exact id accepted by history lookup.
      */
     history_id: string;
-    labels?: HistoryLabels3 | undefined;
+    labels?: HistoryLabels6 | undefined;
     /**
      * The effective top-level model for the run, if any.
      */
@@ -355,6 +647,7 @@ export type HistoryRecord = {
      * every record so a reader can resolve a session by name from any line.
      */
     name: string;
+    observed_tool_ms?: never | undefined;
     /**
      * The normalized approval mode requested for the run.
      */
@@ -394,7 +687,7 @@ export type HistoryRecord = {
      */
     timestamp: string;
     tool_ms?: never | undefined;
-    usage: Usage3;
+    usage: Usage6;
     variant?: string | null | undefined;
     [k: string]: unknown;
 } | {
@@ -424,7 +717,7 @@ export type HistoryRecord = {
      * by `history watch --after` and the exact id accepted by history lookup.
      */
     history_id: string;
-    labels?: HistoryLabels4 | undefined;
+    labels?: HistoryLabels7 | undefined;
     /**
      * The effective top-level model for the run, if any.
      */
@@ -435,6 +728,11 @@ export type HistoryRecord = {
      * every record so a reader can resolve a session by name from any line.
      */
     name: string;
+    /**
+     * Union of tool intervals observed at the stdout pipe. Unlike `tool_ms`,
+     * this is not provider-measured and has no model-latency counterpart.
+     */
+    observed_tool_ms?: number | null | undefined;
     /**
      * The normalized approval mode requested for the run.
      */
@@ -480,11 +778,15 @@ export type HistoryRecord = {
      */
     timestamp: string;
     tool_ms?: number | null | undefined;
-    usage: Usage4;
+    usage: Usage7;
     variant?: string | null | undefined;
     [k: string]: unknown;
 };
 export type ToolCallStatus = "completed" | "failed" | "timeout" | "interrupted";
+/**
+ * How a normalized tool interval was obtained.
+ */
+export type TimingSource = "provider_measured" | "stdout_observed";
 /**
  * The normalized, closed set of failure reasons oneharness can classify from a
  * harness's output. It is the single source for the `failure_kind` contract
@@ -545,6 +847,10 @@ export interface ActionEvent {
      * Terminal tool state, populated on history tool-call events.
      */
     status: ToolCallStatus | null;
+    /**
+     * Provenance for the tool interval. Omitted when timing is unavailable.
+     */
+    timing_source?: TimingSource | null | undefined;
     /**
      * Stable call identity within the session. Present on tool calls and their
      * matching results when the provider exposes an identity; history fills a
@@ -738,3 +1044,114 @@ export interface Usage4 {
     output_tokens: number | null;
     [k: string]: unknown;
 }
+/**
+ * Caller-supplied metadata used to select related task-graph records.
+ * Omitted on the wire when empty for additive compatibility.
+ */
+export interface HistoryLabels5 {
+    [k: string]: string;
+}
+/**
+ * Best-effort token/cost accounting (every field `null` when unreported).
+ */
+export interface Usage5 {
+    /**
+     * Prompt tokens served from the provider's prompt cache (a cheap read of a
+     * previously-written prefix), when the harness reports them. `None` when the
+     * harness does not surface cache counts — never `0` as a guess.
+     */
+    cache_read_tokens: number | null;
+    /**
+     * Prompt tokens written to the provider's prompt cache (a.k.a. cache
+     * creation), when the harness reports them. `None` when not surfaced.
+     */
+    cache_write_tokens: number | null;
+    /**
+     * Total cost in USD, when the harness reports it (often absent on
+     * subscription auth, where there is no per-call dollar figure).
+     */
+    cost_usd: number | null;
+    /**
+     * Prompt/input tokens billed, when the harness reports them.
+     */
+    input_tokens: number | null;
+    /**
+     * Completion/output tokens billed, when the harness reports them.
+     */
+    output_tokens: number | null;
+    [k: string]: unknown;
+}
+/**
+ * Caller-supplied metadata used to select related task-graph records.
+ * Omitted on the wire when empty for additive compatibility.
+ */
+export interface HistoryLabels6 {
+    [k: string]: string;
+}
+/**
+ * Best-effort token/cost accounting (every field `null` when unreported).
+ */
+export interface Usage6 {
+    /**
+     * Prompt tokens served from the provider's prompt cache (a cheap read of a
+     * previously-written prefix), when the harness reports them. `None` when the
+     * harness does not surface cache counts — never `0` as a guess.
+     */
+    cache_read_tokens: number | null;
+    /**
+     * Prompt tokens written to the provider's prompt cache (a.k.a. cache
+     * creation), when the harness reports them. `None` when not surfaced.
+     */
+    cache_write_tokens: number | null;
+    /**
+     * Total cost in USD, when the harness reports it (often absent on
+     * subscription auth, where there is no per-call dollar figure).
+     */
+    cost_usd: number | null;
+    /**
+     * Prompt/input tokens billed, when the harness reports them.
+     */
+    input_tokens: number | null;
+    /**
+     * Completion/output tokens billed, when the harness reports them.
+     */
+    output_tokens: number | null;
+    [k: string]: unknown;
+}
+/**
+ * Caller-supplied metadata used to select related task-graph records.
+ * Omitted on the wire when empty for additive compatibility.
+ */
+export interface HistoryLabels7 {
+    [k: string]: string;
+}
+/**
+ * Best-effort token/cost accounting (every field `null` when unreported).
+ */
+export interface Usage7 {
+    /**
+     * Prompt tokens served from the provider's prompt cache (a cheap read of a
+     * previously-written prefix), when the harness reports them. `None` when the
+     * harness does not surface cache counts — never `0` as a guess.
+     */
+    cache_read_tokens: number | null;
+    /**
+     * Prompt tokens written to the provider's prompt cache (a.k.a. cache
+     * creation), when the harness reports them. `None` when not surfaced.
+     */
+    cache_write_tokens: number | null;
+    /**
+     * Total cost in USD, when the harness reports it (often absent on
+     * subscription auth, where there is no per-call dollar figure).
+     */
+    cost_usd: number | null;
+    /**
+     * Prompt/input tokens billed, when the harness reports them.
+     */
+    input_tokens: number | null;
+    /**
+     * Completion/output tokens billed, when the harness reports them.
+     */
+    output_tokens: number | null;
+    [k: string]: unknown;
+}

package/dist/generated/run-stream-envelope.d.ts

@@ -16,6 +16,10 @@ export type RunStreamEnvelope = {
     [k: string]: unknown;
 };
 export type ToolCallStatus = "completed" | "failed" | "timeout" | "interrupted";
+/**
+ * How a normalized tool interval was obtained.
+ */
+export type TimingSource = "provider_measured" | "stdout_observed";
 /**
  * The normalized, closed set of failure reasons oneharness can classify from a
  * harness's output. It is the single source for the `failure_kind` contract
@@ -84,6 +88,10 @@ export interface ActionEvent {
      * Terminal tool state, populated on history tool-call events.
      */
     status: ToolCallStatus | null;
+    /**
+     * Provenance for the tool interval. Omitted when timing is unavailable.
+     */
+    timing_source?: TimingSource | null | undefined;
     /**
      * Stable call identity within the session. Present on tool calls and their
      * matching results when the provider exposes an identity; history fills a