docs(grep): describe the real output cap

src/fastcontext/agent/tool/grep.md

@@ -6,5 +6,5 @@ Usage:
 - Output modes: "content" shows matching lines (default), "files_with_matches" shows only file paths, "count" shows match counts
 - Pattern syntax: Uses ripgrep (not grep) - literal braces need escaping (use interface\{\} to find interface{} in Go code)
 - Multiline matching: By default patterns match within single lines only. For cross-line patterns like struct \{[\s\S]*?field, use multiline: true
-- Results are capped to several thousand output lines for responsiveness; when truncation occurs, the results report "at least" counts, but are otherwise accurate.
+- Output is capped at the first 100 lines by default for responsiveness; raise or lower this with the head_limit parameter. When truncation occurs, the output ends with a "Results truncated to first N lines" note.
 - Content output formatting closely follows ripgrep output format: '-' for context lines, ':' for match lines, and all context/match lines below each file group.

tests/test_grep_doc_cap.py

@@ -0,0 +1,30 @@
+import asyncio
+import json
+import tempfile
+from pathlib import Path
+
+from fastcontext.agent.tool.grep import GrepTool
+
+
+def test_grep_doc_describes_real_cap():
+    """The Grep description must describe the real 100-line cap and the
+    actual truncation note, not the fictional 'several thousand'/'at least'
+    behavior."""
+    desc = GrepTool().description
+    assert "several thousand" not in desc
+    assert "at least" not in desc
+    assert "100" in desc
+    assert "head_limit" in desc
+    # The documented truncation wording must match what the tool actually emits.
+    assert "Results truncated to first" in desc
+
+
+def test_grep_doc_matches_runtime_truncation():
+    """Cross-check: the truncation note produced at runtime uses exactly the
+    phrasing the doc promises."""
+    grep = GrepTool()
+    with tempfile.TemporaryDirectory() as cwd:
+        (Path(cwd) / "haystack.txt").write_text("\n".join("MATCH" for _ in range(150)), encoding="utf-8")
+        out = asyncio.run(grep.call(json.dumps({"pattern": "MATCH", "output_mode": "content"}), cwd=cwd))
+        assert "Results truncated to first 100 lines" in out
+        assert "Results truncated to first" in GrepTool().description