pi-tool-repair 0.1.4 → 0.1.6

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pi-tool-repair",
-  "version": "0.1.4",
+  "version": "0.1.6",
   "description": "Validate-then-repair extension for pi — fixes common LLM tool-call mistakes (null fields, stringified arrays, wrong field names, anchor bleed) before tools execute",
   "type": "module",
   "author": "Tom X Nguyen",
@@ -33,13 +33,6 @@
     "docs/",
     "README.md"
   ],
-  "scripts": {
-    "test": "vitest run",
-    "test:watch": "vitest",
-    "test:coverage": "vitest run --coverage",
-    "typecheck": "tsc --noEmit",
-    "lint:dead": "knip --no-gitignore"
-  },
   "devDependencies": {
     "@earendil-works/pi-coding-agent": "0.79.4",
     "@types/node": "25.9.1",
@@ -52,5 +45,12 @@
     "extensions": [
       "./tool-repair.ts"
     ]
+  },
+  "scripts": {
+    "test": "vitest run",
+    "test:watch": "vitest",
+    "test:coverage": "vitest run --coverage",
+    "typecheck": "tsc --noEmit",
+    "lint:dead": "knip --no-gitignore"
   }
-}
+}

package/src/index.ts

@@ -131,6 +131,80 @@ export function stripAnchorBleedInPlace(obj: Record<string, unknown>): void {
   }
 }
 
+// Leaked grammar markers from GLM/ChatGLM style tool-call grammars.
+// These can end up as literal prefixes/suffixes on parsed object keys
+// or string values instead of being interpreted as XML tags.
+const GRAMMAR_TOKEN_LEAKS = [
+  { tag: "<arg_key>", at: "start" as const },
+  { tag: "</arg_key>", at: "end" as const },
+  { tag: "<arg_value>", at: "start" as const },
+  { tag: "</arg_value>", at: "end" as const },
+];
+
+export function stripGrammarTokenLeaksInPlace(obj: Record<string, unknown>): boolean {
+  let changed = false;
+  for (const key of Object.keys(obj)) {
+    const value = obj[key];
+    let newKey = key;
+    for (const { tag, at } of GRAMMAR_TOKEN_LEAKS) {
+      if (at === "start" && newKey.startsWith(tag)) {
+        newKey = newKey.slice(tag.length);
+      } else if (at === "end" && newKey.endsWith(tag)) {
+        newKey = newKey.slice(0, -tag.length);
+      }
+    }
+    newKey = newKey.trim();
+
+    if (newKey !== key) {
+      obj[newKey] = value;
+      delete obj[key];
+      changed = true;
+    }
+
+    if (typeof value === "string") {
+      let s = value;
+      for (const { tag, at } of GRAMMAR_TOKEN_LEAKS) {
+        if (at === "start" && s.startsWith(tag)) {
+          s = s.slice(tag.length);
+        } else if (at === "end" && s.endsWith(tag)) {
+          s = s.slice(0, -tag.length);
+        }
+      }
+      const trimmed = s.trim();
+      if (trimmed !== value) {
+        obj[newKey] = trimmed;
+        changed = true;
+      }
+    } else if (Array.isArray(value)) {
+      for (let i = 0; i < value.length; i++) {
+        const item = value[i];
+        if (typeof item === "string") {
+          let s = item;
+          for (const { tag, at } of GRAMMAR_TOKEN_LEAKS) {
+            if (at === "start" && s.startsWith(tag)) {
+              s = s.slice(tag.length);
+            } else if (at === "end" && s.endsWith(tag)) {
+              s = s.slice(0, -tag.length);
+            }
+          }
+          const trimmed = s.trim();
+          if (trimmed !== item) {
+            value[i] = trimmed;
+            changed = true;
+          }
+        } else if (item && typeof item === "object") {
+          const nestedChanged = stripGrammarTokenLeaksInPlace(item as Record<string, unknown>);
+          if (nestedChanged) changed = true;
+        }
+      }
+    } else if (value && typeof value === "object") {
+      const nestedChanged = stripGrammarTokenLeaksInPlace(value as Record<string, unknown>);
+      if (nestedChanged) changed = true;
+    }
+  }
+  return changed;
+}
+
 // ─── Phase 2: Repair Rules ────────────────────────────────────────────────────
 
 export interface RepairResult {

package/tool-repair.ts

@@ -45,6 +45,7 @@ import {
   hasAnchorBleedBug,
   sanitizeSchemaAnchors,
   stripAnchorBleedInPlace,
+  stripGrammarTokenLeaksInPlace,
   wrapRootStringAsObject,
   repairToolInput,
   validateAgainstSchema,
@@ -133,8 +134,31 @@ export default function (pi: ExtensionAPI) {
   pi.on("message_end", (event) => {
     if (event.message.role !== "assistant") return;
 
-    // Phase 1.5: Phantom toolUse normalization (always-on)
     const currentMessage = event.message as unknown as MinimalAssistantMessage;
+
+    // Phase 1.4: Strip leaked grammar tokens from existing pi toolCall blocks.
+    // These can leak from GLM-style grammars straight into the parsed argument
+    // keys (e.g. "<arg_key>command" instead of "command"). We have to repair
+    // them here, before pi preflights and validates the tool calls.
+    let toolCallArgsChanged = false;
+    if (Array.isArray(currentMessage.content)) {
+      for (const part of currentMessage.content) {
+        if (
+          part && typeof part === "object" && !Array.isArray(part) &&
+          (part as Record<string, unknown>).type === "toolCall"
+        ) {
+          const args = (part as Record<string, unknown>).arguments;
+          if (
+            args && typeof args === "object" && !Array.isArray(args) &&
+            stripGrammarTokenLeaksInPlace(args as Record<string, unknown>)
+          ) {
+            toolCallArgsChanged = true;
+          }
+        }
+      }
+    }
+
+    // Phase 1.5: Phantom toolUse normalization (always-on)
     const phantomResult = normalizePhantomToolUse(currentMessage);
 
     if (phantomResult.changed) {
@@ -173,7 +197,10 @@ export default function (pi: ExtensionAPI) {
     }
 
     // Phase 1: Grammar leak repair (opt-in)
-    if (!grammarRepairConfig.enabled) return;
+    if (!grammarRepairConfig.enabled) {
+      if (toolCallArgsChanged) return { message: currentMessage as any };
+      return;
+    }
 
     const knownTools = new Set(
       safeGetActiveTools(pi)
@@ -186,9 +213,9 @@ export default function (pi: ExtensionAPI) {
       knownTools,
     );
 
-    if (!result.changed) return;
+    if (!result.changed && !toolCallArgsChanged) return;
 
-    if (grammarRepairConfig.debug) {
+    if (grammarRepairConfig.debug && result.changed) {
       const calls = result.recoveredCalls.map((call) => `${call.grammar}:${call.name}`).join(",") || "none";
       process.stderr.write(
         `[pi-tool-repair] grammar-repair mode=${grammarRepairConfig.mode} ` +
@@ -213,6 +240,12 @@ export default function (pi: ExtensionAPI) {
       }
     }
 
+    // Defense-in-depth: strip leaked grammar tokens (e.g. GLM <arg_key>) from
+    // parsed tool-call keys/values before the schema sees them.
+    if (input && typeof input === "object") {
+      stripGrammarTokenLeaksInPlace(input as Record<string, unknown>);
+    }
+
     // Only repair built-in tools we have schemas for
     const schema = BUILTIN_SCHEMAS[toolName];
     if (!schema) return;