AI Agents First — Claude Code Dynamic Workflows
Public MindPattern findings, entities, and graph evidence that cite this source.
Findings
5
All-time hits
5
High value
0
Last seen
2026-07-09
Connected entities
Related findings
- 2026-07-09 / SKILLSLet the lead agent plan and fan out to tens–hundreds of subagents — but keep the 3–5 rule for everyday workAnthropic's 2026 Dynamic Workflows lets a lead agent plan and spawn tens to hundreds of parallel subagents in a single session, useful for tasks that genuinely fan out (e.g. sweeping a benchmark across 80 model-prompt combos). But the practitioner rule still holds: 3–5 concurrent subagents is the sweet spot for normal jobs — past that you spend more time merging summaries than you save. Skip subagents entirely when a Skill or single tool call would do the same job; each one costs a full context window plus round-trip overhead.
- 2026-07-07 / SKILLSWire the new agent-lifecycle Notification hooks (agent_needs_input / agent_completed)v2.1.198 shipped two Notification hook events for the background-agent lifecycle: agent_needs_input fires when a delegated agent stalls waiting on you, and agent_completed fires when it finishes. This closes the 'come back 20 minutes later to check' gap for parallel agent fleets. Actionable: hook agent_needs_input to a desktop/Slack ping so blocked agents surface immediately, and agent_completed to trigger your review or CI step.
- 2026-07-07 / SKILLSYour Explore/search subagent got smarter — it now inherits the session model (capped at Opus 4.8)As of v2.1.198 the Explore agent was upgraded from Haiku to inheriting the main session's model (capped at Opus 4.8), so read-only codebase exploration now runs at frontier quality instead of a cheap tier. Delegated search and mapping tasks return materially better results. Actionable: push more broad 'find/where/how does X work' fan-out onto the Explore agent now that it reasons at session strength, and reserve your main context for synthesis.
- 2026-06-29 / SKILLSRun your subagent grader in an isolated context so it can't be gamed by the work that produced the resultAnthropic's Dynamic Workflows (shipped May 28, 2026 with Opus 4.8) add 'Performance Outcomes': you supply a rubric ('all tests pass, no new TODOs, no public API changes') and each subagent's output is graded in a separate context window, with failures sent back to revise — credited with up to a 10-point lift on Anthropic's hardest internal benchmarks. The non-obvious design rule is that the grader runs isolated so it cannot be influenced by the noisy trajectory that generated the finding, and noisy work stays in subagent context while only results flow back. Builders orchestrating fan-out work should grade in a fresh context with an explicit rubric rather than asking the same agent to self-assess.
- 2026-06-26 / SKILLSOrchestrate fan-out with a JS script using agent()/parallel()/pipeline() instead of a chat loopClaude Code Dynamic Workflows let one lead agent write a deterministic JavaScript orchestration script that fans work across up to 1,000 parallel subagents in a single run, using three primitives: agent(), parallel() (barrier), and pipeline() (no barrier between stages). The key win over model-driven dispatch is deterministic control flow — loops, conditionals, and fan-out behave predictably and pipeline() lets item A reach stage 3 while item B is still in stage 1, so wall-clock equals the slowest single chain, not sum-of-stages.