DEV Community

Claude Code subagents, measured Series' Articles

Back to Rulestack's Series
Claude Code subagents: when to delegate, and when a rule or hook is enough

Claude Code subagents: when to delegate, and when a rule or hook is enough

2
Comments 2
7 min read
Claude Code subagents: how .claude/agents files work, and why Claude never uses yours

Claude Code subagents: how .claude/agents files work, and why Claude never uses yours

Comments
4 min read
What a Claude Code subagent actually costs: measuring the ~436k-token fixed overhead

What a Claude Code subagent actually costs: measuring the ~436k-token fixed overhead

Comments 2
5 min read
Your CLAUDE.md loads into every subagent — the context multiplier nobody budgets for

Your CLAUDE.md loads into every subagent — the context multiplier nobody budgets for

Comments 11
5 min read
Claude Code fork vs subagent: what /fork, /subtask, and subagent_type: "fork" each copy

Claude Code fork vs subagent: what /fork, /subtask, and subagent_type: "fork" each copy

Comments 8
5 min read
A Claude Code subagent costs ~436k tokens before it reads a single file: measured, with the break-even math

A Claude Code subagent costs ~436k tokens before it reads a single file: measured, with the break-even math

Comments 4
5 min read
Three Claude Code subagent files, three models in frontmatter: what each launch cost, and the field the validator waved through

Three Claude Code subagent files, three models in frontmatter: what each launch cost, and the field the validator waved through

2
Comments 2
10 min read
A Claude Code subagent wrote 25 posts into our working tree. The parent's git rebase --skip erased all 25

A Claude Code subagent wrote 25 posts into our working tree. The parent's git rebase --skip erased all 25

Comments
10 min read
Bash(git push *) in a Claude Code subagent's disallowedTools took away all of Bash in 10 of 10 launches, not just git push

Bash(git push *) in a Claude Code subagent's disallowedTools took away all of Bash in 10 of 10 launches, not just git push

Comments 1
15 min read
89 of 190 reviewer subagents graded their own model's work. A PreToolUse hook has blocked 7 since; one walked around it

89 of 190 reviewer subagents graded their own model's work. A PreToolUse hook has blocked 7 since; one walked around it

Comments 2
11 min read
One Claude Code session ran 15 subagents under our 10-subagent rule. After a PreToolUse hook, 676 launches peaked at 10

One Claude Code session ran 15 subagents under our 10-subagent rule. After a PreToolUse hook, 676 launches peaked at 10

Comments 2
11 min read