open-agents-ai 0.185.61 → 0.185.62

package/dist/index.js

@@ -26166,11 +26166,16 @@ Respond with your assessment, then take action.`;
           content: `Context assembled: ${contextComposition.sections.map((s) => `${s.label}(${s.tokenEstimate}t)`).join(" + ")} = ~${contextComposition.totalTokenEstimate}t`,
           timestamp: (/* @__PURE__ */ new Date()).toISOString()
         });
+        const MATH_SIGNALS = /\$[\d,]+|\d+%|\bpercent\b|\bcalculate\b|\bcompute\b|\baverage\b|\btotal\b|\bsum\b|\bratio\b|\bconvert\b.*\b(?:to|into)\b|\d+\s*[\+\-\*\/]\s*\d+/i;
+        let userContent = context ? `${context}
+
+TASK: ${task}` : task;
+        if (MATH_SIGNALS.test(task)) {
+          userContent += "\n\n[Note: This involves numerical computation. Use repl_exec or shell to execute Python for all arithmetic \u2014 do not compute in your head.]";
+        }
         const messages = [
           { role: "system", content: systemPrompt },
-          { role: "user", content: context ? `${context}
-
-TASK: ${task}` : task }
+          { role: "user", content: userContent }
         ];
         let toolDefs = this.buildToolDefinitions();
         let textToolModeActive = this.options.textToolMode ?? false;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "open-agents-ai",
-  "version": "0.185.61",
+  "version": "0.185.62",
   "description": "AI coding agent powered by open-source models (Ollama/vLLM) — interactive TUI with agentic tool-calling loop",
   "type": "module",
   "main": "./dist/index.js",

package/prompts/agentic/system-medium.md

@@ -91,6 +91,19 @@ You are **Open Agent** (open-agents-ai), an autonomous AI coding agent running o
 
 When asked "how do you work?" or "what can you do?", answer from this list and use explore_tools() or skill_list() to provide specifics. Do NOT hallucinate capabilities — use tools to discover concrete information.
 
+## Calculations — Always Execute, Never Guess
+
+For ANY numerical calculation involving 2+ operations, write Python and execute it with `repl_exec` or `shell`. In-head arithmetic is error-prone across all model sizes. Python is exact.
+
+```
+User: What is 15% of $847.50 after a $50 discount?
+You: repl_exec(code="result = (847.50 - 50) * 0.15; print(f'${result:.2f}')")
+Output: $119.63
+Answer: $119.63
+```
+
+This applies to: currency conversion, percentages, statistics, financial calculations, unit conversions, date math. If code execution fails, reason through the expected output step by step and mark with [ESTIMATED].
+
 ## Debugging — Observe Before Reasoning
 
 When uncertain about runtime behavior (types, return values, edge cases), run a quick test instead of guessing:

package/prompts/agentic/system-small.md

@@ -28,6 +28,10 @@ Rules:
 - Memory: your persistent memories live in .oa/memory/ — use memory_read(topic) to recall, memory_write(topic, key, value) to save. Session history: file_read(".oa/context/session-diary.md")
 - When asked "what can you do?", use explore_tools() and skill_list() to discover and report your actual capabilities. Do NOT hallucinate.
 
+Calculations — EXECUTE, never guess:
+- For ANY math with 2+ operations: use `repl_exec(code="print(847.50 * 0.15)")` or `shell`. Python is exact. In-head arithmetic is not.
+- Currency, percentages, statistics, dates — ALWAYS execute code. If execution fails, reason step-by-step and mark [ESTIMATED].
+
 Debugging — OBSERVE before reasoning:
 - When unsure how code behaves at runtime, DO NOT guess. Write a short test script and RUN it:
   shell(command="node -e \"console.log(JSON.parse(JSON.stringify({d: new Date()})))\"")