pi-langfuse 1.5.6 → 1.5.7

package/README.md

@@ -104,6 +104,27 @@ export LANGFUSE_CAPTURE_CWD=false
 
 All captured payloads are redacted before upload. The extension masks common API keys, bearer tokens, passwords, cookies, private keys, Langfuse keys, GitHub/npm/AWS-style tokens, and local absolute paths.
 
+### Payload limits
+
+Before upload, payloads are shaped: strings are truncated and deeply nested or
+very wide structures are trimmed. These caps keep traces small and protect the
+Langfuse ingestion pipeline. Override any of them (no rebuild needed):
+
+```bash
+export PI_LANGFUSE_MAX_STRING_LENGTH=12000       # per-string chars (system prompt, inputs)
+export PI_LANGFUSE_MAX_TOOL_PAYLOAD_LENGTH=24000 # per tool input/output chars
+export PI_LANGFUSE_MAX_DEPTH=6                    # max nesting depth
+export PI_LANGFUSE_MAX_ARRAY_ITEMS=50            # max array elements kept
+export PI_LANGFUSE_MAX_OBJECT_KEYS=80            # max object keys kept
+export PI_LANGFUSE_MAX_PAYLOAD_NODES=2000        # max total nodes per payload
+```
+
+Set any limit to `0`, `off`, `none`, or `unlimited` to disable that cap
+entirely (captures the full value). Unset or invalid values fall back to the
+defaults shown above. To capture a very large system prompt or big tool
+payloads in full, raise or disable the relevant limit (e.g.
+`PI_LANGFUSE_MAX_STRING_LENGTH=off`).
+
 ### Method 3: Persistent `config.json`
 
 Create or update `~/.pi/agent/pi-langfuse/config.json`:

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "pi-langfuse",
-	"version": "1.5.6",
+	"version": "1.5.7",
 	"description": "Langfuse extension for Pi coding agent",
 	"repository": {
 		"type": "git",

package/src/config.ts

@@ -5,6 +5,7 @@ import { CONFIG_PATH, DEFAULT_LANGFUSE_HOST } from "./constants.js";
 import { state } from "./state.js";
 import { forceShutdownRuntime } from "./langfuse.js";
 import { createCapturePolicy, type EnvLike } from "./capture-policy.js";
+import { createPayloadLimits } from "./limits.js";
 
 export function loadConfigFromFile(path = CONFIG_PATH, env: EnvLike = process.env as EnvLike): Config | null {
   if (existsSync(path)) {
@@ -22,6 +23,7 @@ export function loadConfigFromFile(path = CONFIG_PATH, env: EnvLike = process.en
           secretKey: config.secretKey,
           host: config.host || DEFAULT_LANGFUSE_HOST,
           capturePolicy: createCapturePolicy(captureSource),
+          limits: createPayloadLimits(env),
         };
       }
     } catch (e) {
@@ -44,6 +46,7 @@ export function loadConfigFromEnv(env: EnvLike = process.env as EnvLike): Config
     secretKey,
     host: env.LANGFUSE_BASE_URL || env.LANGFUSE_HOST || DEFAULT_LANGFUSE_HOST,
     capturePolicy: createCapturePolicy(env),
+    limits: createPayloadLimits(env),
   };
 }
 

package/src/handlers/agent.ts

@@ -1,7 +1,7 @@
 import { state, resetRunState, computeEvaluationScores } from "../state.js";
 import { getRuntime, sendScore } from "../langfuse.js";
 import { ensureConfig } from "../config.js";
-import { shapePayload, truncate, extractFinalAssistant, extractAssistantOutput, getCapturePolicy } from "../utils.js";
+import { shapePayload, truncate, extractFinalAssistant, extractAssistantOutput, getCapturePolicy, getLimits } from "../utils.js";
 import { closeDanglingObservations } from "./tool.js";
 import { applyCapturePolicy } from "../capture-policy.js";
 import { collectSourceMetadata } from "../source-metadata.js";
@@ -79,7 +79,7 @@ export async function startAgentRun(event: Record<string, unknown>, ctx: any) {
           ...(state.currentProvider ? { provider: state.currentProvider } : {}),
           sessionId: state.currentSessionId || undefined,
         },
-        systemPrompt: systemPrompt ? truncate(String(systemPrompt), 20000) : undefined,
+        systemPrompt: systemPrompt ? truncate(String(systemPrompt), getLimits().maxString) : undefined,
       },
       getCapturePolicy(),
     );

package/src/handlers/tool.ts

@@ -10,8 +10,8 @@ import {
   truncate,
   estimatePayloadBytes,
   getCapturePolicy,
+  getLimits,
 } from "../utils.js";
-import { MAX_TOOL_PAYLOAD_LENGTH } from "../constants.js";
 import { applyCapturePolicy } from "../capture-policy.js";
 import { redactString } from "../redaction.js";
 
@@ -28,7 +28,7 @@ export async function startToolObservation(event: Record<string, unknown>) {
   try {
     const toolName = getToolName(event);
     const toolInput = getToolInput(event);
-    const shapedInput = shapePayload(toolInput, { maxString: MAX_TOOL_PAYLOAD_LENGTH });
+    const shapedInput = shapePayload(toolInput, { maxString: getLimits().maxToolPayload });
     const captured = applyCapturePolicy(
       {
         toolInput: shapedInput,
@@ -36,7 +36,7 @@ export async function startToolObservation(event: Record<string, unknown>) {
       },
       getCapturePolicy(),
     );
-    const inputBytes = estimatePayloadBytes(captured.toolInput, MAX_TOOL_PAYLOAD_LENGTH);
+    const inputBytes = estimatePayloadBytes(captured.toolInput, getLimits().maxToolPayload);
     const parent = state.agentState.activeTurn ?? state.agentState.root;
     const tool = await startChildObservation({
       parent,
@@ -79,7 +79,7 @@ export async function finishToolObservation(event: Record<string, unknown>) {
 
   const isError = Boolean(event.isError ?? event.error ?? event.status === "error");
   const output =
-    extractTextContent(event.content, MAX_TOOL_PAYLOAD_LENGTH) ??
+    extractTextContent(event.content, getLimits().maxToolPayload) ??
     event.output ??
     event.result ??
     event.error ??
@@ -87,7 +87,7 @@ export async function finishToolObservation(event: Record<string, unknown>) {
     event;
 
   try {
-    const shapedOutput = shapePayload(output, { maxString: MAX_TOOL_PAYLOAD_LENGTH });
+    const shapedOutput = shapePayload(output, { maxString: getLimits().maxToolPayload });
     const captured = applyCapturePolicy(
       {
         toolOutput: shapedOutput,
@@ -99,7 +99,7 @@ export async function finishToolObservation(event: Record<string, unknown>) {
       },
       getCapturePolicy(),
     );
-    const outputBytes = estimatePayloadBytes(captured.toolOutput, MAX_TOOL_PAYLOAD_LENGTH);
+    const outputBytes = estimatePayloadBytes(captured.toolOutput, getLimits().maxToolPayload);
     const durationMs = Math.max(0, Date.now() - activeTool.startedAt);
 
     activeTool.observation

package/src/limits.ts

@@ -0,0 +1,97 @@
+import {
+  MAX_ARRAY_ITEMS,
+  MAX_DEPTH,
+  MAX_OBJECT_KEYS,
+  MAX_PAYLOAD_NODES,
+  MAX_STRING_LENGTH,
+  MAX_TOOL_PAYLOAD_LENGTH,
+} from "./constants.js";
+import type { EnvLike } from "./capture-policy.js";
+import { state } from "./state.js";
+
+/**
+ * Payload-shaping limits. Every field is a positive integer, or
+ * `Number.POSITIVE_INFINITY` to disable that limit entirely (capture everything).
+ * Resolved once from the environment and stored on the loaded config; consumers
+ * read the resolved values via `getLimits()` rather than the raw constants.
+ */
+export interface PayloadLimits {
+  /** Max characters kept per captured string (generation/agent inputs, outputs, system prompt). */
+  readonly maxString: number;
+  /** Max characters kept for tool inputs/outputs (their payloads run larger than chat strings). */
+  readonly maxToolPayload: number;
+  /** Max nesting depth walked when shaping a structured payload. */
+  readonly maxDepth: number;
+  /** Max array elements kept per array. */
+  readonly maxArrayItems: number;
+  /** Max own-keys kept per object. */
+  readonly maxObjectKeys: number;
+  /** Max total nodes visited across a whole payload before bailing with `[payload too large]`. */
+  readonly maxNodes: number;
+}
+
+export const DEFAULT_LIMITS: PayloadLimits = {
+  maxString: MAX_STRING_LENGTH,
+  maxToolPayload: MAX_TOOL_PAYLOAD_LENGTH,
+  maxDepth: MAX_DEPTH,
+  maxArrayItems: MAX_ARRAY_ITEMS,
+  maxObjectKeys: MAX_OBJECT_KEYS,
+  maxNodes: MAX_PAYLOAD_NODES,
+};
+
+/** Words that mean "no limit" when supplied as an env value. */
+const UNLIMITED_WORDS = new Set(["off", "none", "false", "no", "unlimited", "inf", "infinity"]);
+
+/**
+ * Parse one limit env value.
+ * - unset / blank / unparseable  -> `fallback` (the built-in default)
+ * - "off"/"none"/"unlimited"/... or a value <= 0 -> `Infinity` (limit removed)
+ * - a positive number -> that integer
+ */
+export function parseLimit(raw: string | undefined, fallback: number): number {
+  if (raw === undefined) {
+    return fallback;
+  }
+  const trimmed = raw.trim().toLowerCase();
+  if (trimmed === "") {
+    return fallback;
+  }
+  if (UNLIMITED_WORDS.has(trimmed)) {
+    return Number.POSITIVE_INFINITY;
+  }
+  const parsed = Number(trimmed);
+  if (!Number.isFinite(parsed)) {
+    return fallback;
+  }
+  if (parsed <= 0) {
+    return Number.POSITIVE_INFINITY;
+  }
+  return Math.floor(parsed);
+}
+
+/**
+ * Resolve payload limits from the environment. Each `PI_LANGFUSE_MAX_*` var overrides
+ * the corresponding default; set any to `0`/`off`/`unlimited` to remove that limit.
+ * Namespaced `PI_LANGFUSE_*` (not `LANGFUSE_*`) to avoid clashing with Langfuse
+ * server env vars such as `LANGFUSE_MAX_EVENT_SIZE_BYTES`.
+ */
+export function createPayloadLimits(env: EnvLike = process.env as EnvLike): PayloadLimits {
+  return {
+    maxString: parseLimit(env.PI_LANGFUSE_MAX_STRING_LENGTH, DEFAULT_LIMITS.maxString),
+    maxToolPayload: parseLimit(env.PI_LANGFUSE_MAX_TOOL_PAYLOAD_LENGTH, DEFAULT_LIMITS.maxToolPayload),
+    maxDepth: parseLimit(env.PI_LANGFUSE_MAX_DEPTH, DEFAULT_LIMITS.maxDepth),
+    maxArrayItems: parseLimit(env.PI_LANGFUSE_MAX_ARRAY_ITEMS, DEFAULT_LIMITS.maxArrayItems),
+    maxObjectKeys: parseLimit(env.PI_LANGFUSE_MAX_OBJECT_KEYS, DEFAULT_LIMITS.maxObjectKeys),
+    maxNodes: parseLimit(env.PI_LANGFUSE_MAX_PAYLOAD_NODES, DEFAULT_LIMITS.maxNodes),
+  };
+}
+
+/**
+ * Resolved limits for the current session: the config-loaded values when a
+ * config is active, otherwise a fresh resolve from the environment. Every
+ * capture/redaction path reads limits through this so a single env change
+ * (or config) governs truncation everywhere.
+ */
+export function getLimits(): PayloadLimits {
+  return state.config?.limits ?? createPayloadLimits();
+}

package/src/redaction.ts

@@ -1,4 +1,5 @@
 import { createHash } from "node:crypto";
+import { getLimits } from "./limits.js";
 
 export const REDACTED = "[REDACTED_SECRET]";
 
@@ -9,12 +10,15 @@ export interface RedactOptions {
   maxStringLength: number;
 }
 
-const DEFAULT_OPTIONS: RedactOptions = {
-  maxDepth: 6,
-  maxArrayItems: 50,
-  maxObjectKeys: 80,
-  maxStringLength: 12_000,
-};
+function defaultOptions(): RedactOptions {
+  const limits = getLimits();
+  return {
+    maxDepth: limits.maxDepth,
+    maxArrayItems: limits.maxArrayItems,
+    maxObjectKeys: limits.maxObjectKeys,
+    maxStringLength: limits.maxString,
+  };
+}
 
 const SECRET_ASSIGNMENT_RE =
   /\b([A-Z0-9_]*(?:SECRET|TOKEN|PASSWORD|PASS|API[_-]?KEY|PRIVATE[_-]?KEY|AUTH|COOKIE)[A-Z0-9_]*)\s*=\s*([^\s"'`]+)/gi;
@@ -36,7 +40,7 @@ function truncate(value: string, maxStringLength: number): string {
 }
 
 export function redactString(value: string, options: Partial<RedactOptions> = {}): string {
-  const merged = { ...DEFAULT_OPTIONS, ...options };
+  const merged = { ...defaultOptions(), ...options };
   const truncated = truncate(value, merged.maxStringLength);
   return truncated
     .replace(PRIVATE_KEY_RE, REDACTED)
@@ -110,6 +114,6 @@ function visit(value: unknown, options: RedactOptions, depth: number, seen: Weak
 }
 
 export function redactValue(value: unknown, options: Partial<RedactOptions> = {}): unknown {
-  const merged: RedactOptions = { ...DEFAULT_OPTIONS, ...options };
+  const merged: RedactOptions = { ...defaultOptions(), ...options };
   return visit(value, merged, merged.maxDepth, new WeakSet<object>());
 }

package/src/types.ts

@@ -1,10 +1,12 @@
 import type { CapturePolicy } from "./capture-policy.js";
+import type { PayloadLimits } from "./limits.js";
 
 export interface Config {
   publicKey: string;
   secretKey: string;
   host: string;
   capturePolicy?: CapturePolicy;
+  limits?: PayloadLimits;
 }
 
 export interface LangfuseObservation {

package/src/utils.ts

@@ -1,11 +1,4 @@
-import {
-  MAX_ARRAY_ITEMS,
-  MAX_DEPTH,
-  MAX_OBJECT_KEYS,
-  MAX_PAYLOAD_NODES,
-  MAX_STRING_LENGTH,
-  MAX_TOOL_PAYLOAD_LENGTH,
-} from "./constants.js";
+import { getLimits } from "./limits.js";
 import { createCapturePolicy, type CapturePolicy } from "./capture-policy.js";
 import { redactValue } from "./redaction.js";
 import { state } from "./state.js";
@@ -14,7 +7,9 @@ export function getCapturePolicy(): CapturePolicy {
   return state.config?.capturePolicy ?? createCapturePolicy();
 }
 
-export function truncate(value: string, maxLength = MAX_STRING_LENGTH): string {
+export { getLimits };
+
+export function truncate(value: string, maxLength = getLimits().maxString): string {
   return value.length > maxLength ? `${value.slice(0, maxLength)}... [truncated]` : value;
 }
 
@@ -35,11 +30,22 @@ const PAYLOAD_TOO_LARGE = "[payload too large]";
 
 export function shapePayload(
   value: unknown,
-  options: { maxString?: number; depth?: number; maxNodes?: number; redact?: boolean; parseJson?: boolean } = {},
+  options: {
+    maxString?: number;
+    depth?: number;
+    maxNodes?: number;
+    maxArrayItems?: number;
+    maxObjectKeys?: number;
+    redact?: boolean;
+    parseJson?: boolean;
+  } = {},
 ): unknown {
-  const maxString = options.maxString ?? MAX_STRING_LENGTH;
-  const depth = options.depth ?? MAX_DEPTH;
-  const maxNodes = options.maxNodes ?? MAX_PAYLOAD_NODES;
+  const limits = getLimits();
+  const maxString = options.maxString ?? limits.maxString;
+  const depth = options.depth ?? limits.maxDepth;
+  const maxNodes = options.maxNodes ?? limits.maxNodes;
+  const maxArrayItems = options.maxArrayItems ?? limits.maxArrayItems;
+  const maxObjectKeys = options.maxObjectKeys ?? limits.maxObjectKeys;
   const budget = { exhausted: false, nodeCount: 0 };
 
   function visit(item: unknown, remainingDepth: number, seen: WeakSet<object>): unknown {
@@ -88,7 +94,7 @@ export function shapePayload(
 
     if (Array.isArray(item)) {
       const output: unknown[] = [];
-      const limit = Math.min(item.length, MAX_ARRAY_ITEMS);
+      const limit = Math.min(item.length, maxArrayItems);
       for (let index = 0; index < limit; index++) {
         output.push(visit(item[index], remainingDepth - 1, seen));
         if (budget.exhausted) {
@@ -120,7 +126,7 @@ export function shapePayload(
         }
         output[key] = visit((item as Record<string, unknown>)[key], remainingDepth - 1, seen);
         keyCount++;
-        if (budget.exhausted || keyCount >= MAX_OBJECT_KEYS) {
+        if (budget.exhausted || keyCount >= maxObjectKeys) {
           break;
         }
       }
@@ -136,12 +142,12 @@ export function shapePayload(
     : redactValue(shaped, {
         maxDepth: depth,
         maxStringLength: maxString,
-        maxArrayItems: MAX_ARRAY_ITEMS,
-        maxObjectKeys: MAX_OBJECT_KEYS,
+        maxArrayItems,
+        maxObjectKeys,
       });
 }
 
-export function safeSerialize(value: unknown, maxLength = MAX_TOOL_PAYLOAD_LENGTH): string {
+export function safeSerialize(value: unknown, maxLength = getLimits().maxToolPayload): string {
   try {
     return truncate(JSON.stringify(shapePayload(value, { maxString: maxLength }), null, 2), maxLength);
   } catch {
@@ -149,7 +155,7 @@ export function safeSerialize(value: unknown, maxLength = MAX_TOOL_PAYLOAD_LENGT
   }
 }
 
-export function estimatePayloadBytes(value: unknown, maxLength = MAX_TOOL_PAYLOAD_LENGTH): number {
+export function estimatePayloadBytes(value: unknown, maxLength = getLimits().maxToolPayload): number {
   return new TextEncoder().encode(safeSerialize(value, maxLength)).length;
 }