What are the most popular Claude Code skills?

Ranked by real invocations measured from developers' local agent transcripts, the skills at the top of this page are the ones developers actually run — not the ones with the most GitHub stars or directory listings. The ranking updates as new usage syncs in.

Which MCP servers do developers actually use?

The tool-call table on this page separates MCP server calls (named like server · tool) from built-in agent tools, ranked by real call counts, token burn and error rates measured from local transcripts.

How is skill and tool usage measured?

The whoburnedmore CLI parses the transcripts coding agents keep locally and submits only name + count + token aggregates for skills and tool calls. No prompts, arguments, file names or outputs are collected.

Most-Used Claude Code Skills & MCP Servers — Real Usage Data

SKILLS+MCP

LIVE · 2026-11-13 06:56Z

Skills tracked

100

Skill invocations

934,929

MCP tools

Built-in tools

Tool calls

9,970,648

Developers

176

listed panel

Skills by real usage

100 skills · row fill = vs top skill

#	Skill	Devs	Calls	Tokens
01	superpowers:writing-plans	31	33,134	9.04B
02	code-review	58	55,278	8.64B
03	workflow-authoring	24	54,434	7.91B

MCP server tools

49 tools · server · tool

#	MCP tool	Devs	Calls	Tokens	Err
01	claude-in-chrome · browser_batch	78	33,807	9.64B	5%
02	ccd_turn · update_plan	66	59,355	9.44B

METHODOLOGY — measured, not surveyed: the open-source npx whoburnedmore CLI reads the local usage logs coding agents already keep and submits daily aggregate token totals per tool and model — never prompts, code or file names. Panel = 176 listed developers who opted in publicly, including subscription usage (Claude Code, Cursor, Codex) that API-gateway rankings never see. Costs estimated from public per-token pricing. what we collect · leaderboard · state of AI coding

Built-in agent tools

51 tools

#	Tool	Devs	Calls	Tokens	Err
01	Bash	280	4,750,257	982.86B	3%
02	exec	206	1,213,583	157.04B	0%
03	Read	280	856,092	118.13B	1%
04	Edit	273	447,247	105.03B	2%
05	Write	275	248,648	57.28B	2%
06	exec_command	122	667,924	36.86B	0%
07	WebFetch	249	184,096	17.53B	4%
08	apply_patch	131	114,707	14.46B	0%
09	Agent	237	55,380	13.34B	2%
10	WebSearch	239	117,350	10.35B	1%
11	send_message	102	74,629	9.72B	0%
12	ToolSearch	254	71,114	9.67B	0%
13	write_stdin	104	91,196	9.51B	0%
14	wait	172	62,940	8.54B	0%
15	js	146	53,953	7.56B	0%
16	shell_command	50	138,328	7.36B	0%
17	PowerShell	49	25,881	6.24B	12%
18	StructuredOutput	95	49,669	5.27B	9%
19	SendMessage	143	21,970	5.16B	0%
20	TaskUpdate	124	18,591	4.73B	0%
21	Monitor	185	16,733	4.56B	4%
22	Grep	121	47,189	4.45B	2%
23	SubagentHandback	113	21,281	4.24B	0%
24	AskUserQuestion	255	13,719	3.40B	7%
25	wait_agent	96	53,787	3.09B	0%
26	Artifact	148	10,272	2.85B	5%
27	Skill	242	18,550	2.79B	1%
28	TaskCreate	126	10,966	2.34B	1%
29	TaskStop	209	9,635	2.25B	4%
30	sleep	100	12,047	1.81B	0%
31	shell	10	21,968	1.76B	0%
32	view_file	2	4,692	1.57B	0%
33	SendUserFile	116	4,567	1.53B	2%
34	update_plan	105	13,290	1.48B	0%
35	manage_task	2	2,239	1.36B	0%
36	run_command	3	5,971	1.05B	0%
37	ScheduleWakeup	125	3,234	1.02B	2%
38	Glob	98	24,682	875.8M	1%
39	Workflow	84	2,552	856.0M	6%
40	followup_task	82	8,077	834.6M	0%
41	TodoWrite	33	3,855	790.0M	0%
42	spawn_agent	118	11,018	685.8M	0%
43	write_to_file	2	2,717	629.8M	0%
44	list_agents	87	6,225	599.2M	0%
45	view_image	86	6,561	567.5M	0%
46	replace_file_content	2	1,392	526.8M	0%
47	ArtifactData	23	2,259	411.7M	14%
48	run	17	3,324	377.9M	0%
49	request_user_input_async	119	2,820	370.1M	0%
50	ListAgents	97	1,354	342.2M	0%
51	bash	65	2,904	314.8M	31%

SKILLS+MCP

▮Skills by real usage

▮MCP server tools

▮Built-in agent tools

Skills by real usage

MCP server tools

Built-in agent tools