• There are 37 files with 8,183 lines of code.
• 1 very long files (1,113 lines of code)
• 5 long files (3,319 lines of code)
• 7 medium size files (2,202 lines of code)
• 7 small files (1,011 lines of code)
• 17 very small files (538 lines of code)

File	# lines	≈ tokens to read	# changes (1y)	≈ lines edited per change	lines read per line changed
types.py in src/claude_agent_sdk	894	~5k–15k	84	~30	~30
subprocess_cli.py in src/claude_agent_sdk/_internal/transport	731	~5k–15k	52	~30	~20
query.py in src/claude_agent_sdk/_internal	678	~3k–10k	41	~50	~15
__init__.py in src/claude_agent_sdk	508	~3k–10k	43	~30	~20
sessions.py in src/claude_agent_sdk/_internal	1113	~7k–20k	10	~200	~5
message_parser.py in src/claude_agent_sdk/_internal	340	~2k–5k	22	~20	~20
client.py in src/claude_agent_sdk	239	~1.5k–5k	28	~30	~10
client.py in src/claude_agent_sdk/_internal	149	~1k–3k	19	~20	~7
session_mutations.py in src/claude_agent_sdk/_internal	508	~3k–10k	5	~200	~2
build_wheel.py in scripts	489	~3k–7k	4	~200	~3
session_resume.py in src/claude_agent_sdk/_internal	380	~2k–7k	5	~150	~3
download_cli.py in scripts	200	~1k–3k	4	~150	~2
session_store_conformance.py in src/claude_agent_sdk/testing	221	~1.5k–5k	3	~100	~2
transcript_mirror_batcher.py in src/claude_agent_sdk/_internal	132	~700–2k	4	~70	~2
check_pypi_quota.py in scripts	102	~700–1.5k	3	~50	~2
check_workflow_hardening.py in .github/scripts	300	~2k–5k	1	~300	~1
session_summary.py in src/claude_agent_sdk/_internal	142	~700–2k	2	~100	~1
_cli_version.py in src/claude_agent_sdk	1	under 1k	256	~2	~0
session_store.py in src/claude_agent_sdk/_internal	119	~700–2k	2	~100	~1
sdk_mcp_bridge.py in src/claude_agent_sdk/_internal	233	~1.5k–5k	1	~300	~1

Tokens are lines × 7 to 14 (analysis.fileReadsForChanges). Lines edited per change are the file's lines added and deleted divided by its commits over the whole history. Agents read about 2,000 lines at a time: when a 17,000-line file was split into modules, the tokens an agent read for the same change fell by 83% (The Economic Benefit of Refactoring, 2026).

There are 3 files with lines longer than 120 characters. In total, there are 5 long lines.

File	# lines	# units	# long lines
run-e2e-firewalled-windows.ps1 in .github/scripts	167	-	3
check_workflow_hardening.py in .github/scripts	300	14	1
query.py in src/claude_agent_sdk/_internal	678	20	1

• File size measurements show the distribution of size of files.
• Files are classified in four categories based on their size (lines of code): 1-100 (very small files), 101-200 (small files), 201-500 (medium size files), 501-1000 (long files), 1001+(very long files).
• It is a good practice to keep files small. Long files may become "bloaters", code that have increased to such gargantuan proportions that they are hard to work with.