Python: Fix MangenticAgentExecutor producing repr str for tool call content

[moonbox3]

Motivation and Context

When an agent's response contains only FunctionCallContent (tool calls) without text content, ChatMessage.text returns an empty string. The code text = last.text or str(last) evaluated the fallback str(last), which produced garbage repr output like <agent_framework._types.ChatMessage object at 0x...>.

Changed to text = last.text or "" to match the existing pattern used elsewhere in the SDK. The contents list still contains the FunctionCallContent items for consumers who need them.

Description

• Fixes Python: MagenticAgentExecutor produces repr string instead of message text for tool-call-only responses #2562

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the Contribution Guidelines
• All unit tests pass, and I have added new tests where possible
• Is this a breaking change? If yes, add "[BREAKING]" prefix to the title of the PR.

[markwallace-microsoft]

Python Test Coverage Report •

File	Stmts	Miss	Cover	Missing
packages/core/agent_framework/_workflows
_magentic.py	859	253	70%	46, 51, 73–82, 87, 91–102, 275, 280, 297, 299, 314, 322–331, 409, 413, 427, 433, 448, 528, 541, 558, 567–568, 570–572, 574, 585, 727–730, 733–737, 739–741, 748, 787, 834, 870–872, 874, 882–885, 889–892, 1022–1023, 1040, 1042–1043, 1051, 1087, 1096–1098, 1118, 1171, 1191, 1194, 1223, 1226, 1234–1238, 1244, 1272–1274, 1276, 1278, 1286–1289, 1291, 1295–1296, 1299–1302, 1304–1305, 1311–1313, 1316–1317, 1322–1323, 1331, 1339, 1354, 1366, 1378–1381, 1410–1411, 1416–1418, 1449, 1475, 1489, 1505, 1521, 1588, 1595, 1598, 1600–1601, 1604–1605, 1609, 1612, 1633, 1668–1669, 1671, 1675–1677, 1735, 1740–1741, 2067–2068, 2072, 2087, 2092, 2095, 2149, 2160, 2171–2173, 2186–2187, 2192, 2203–2205, 2216–2218, 2230–2237, 2239–2240, 2248, 2256–2257, 2259–2261, 2263–2266, 2270–2278, 2282–2283, 2286–2290, 2292–2293, 2295–2298, 2300–2303, 2305–2307, 2309–2310, 2312–2314, 2329–2332, 2343–2346, 2358–2361, 2365
TOTAL	15873	2412	84%

Python Unit Test Overview

Tests	Skipped	Failures	Errors	Time
2278	130 💤	0 ❌	0 🔥	54.225s ⏱️

[Copilot]

Pull request overview

This PR fixes a bug where MagenticAgentExecutor was producing garbage repr output when agent responses contained only tool calls (FunctionCallContent) without text content. The fix replaces str(last) fallback with an empty string "", aligning with the established pattern used throughout the codebase.

Key Changes:

• Changed fallback from str(last) to "" when ChatMessage.text is empty
• Applied the fix consistently in both the core workflow and DevUI mapper

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
python/packages/core/agent_framework/_workflows/_magentic.py	Fixed _invoke_agent method to use empty string instead of repr output when message has no text content
python/packages/devui/agent_framework_devui/_mapper.py	Fixed _convert_workflow_event method to use empty string instead of repr output for orchestrator messages without text

---

You can also share your feedback on Copilot code review for a chance to win a $100 gift card. Take the survey.