dispatching-parallel-agents

Name: dispatching-parallel-agents
Author: obra

byobra· 95 skills

Installs

149k

Stars

261k

Forks

23k

How to install dispatching-parallel-agents?

obra/superpowers/dispatching-parallel-agents

$npx -y skills add obra/superpowers --skill dispatching-parallel-agents

Installs into the current project.

›Prefer a prompt? Paste this to your agent

Use this skill

Run `npx skills use "https://github.com/obra/superpowers" --skill "obra/superpowers/dispatching-parallel-agents"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.

Use the whole pack

Use the skills in "https://github.com/obra/superpowers" that are relevant to the current task. Run `npx skills add "https://github.com/obra/superpowers"` and select the relevant skills, then follow their instructions.

Files · 1

View on GitHub

SKILL.md

1	# Dispatching Parallel Agents
2
3	## Overview
4
5	You delegate tasks to specialized agents with isolated context. By precisely crafting their instructions and context, you ensure they stay focused and succeed at their task. They should never inherit your session's context or history — you construct exactly what they need. This also preserves your own context for coordination work.
6
7	When you have multiple unrelated failures (different test files, different subsystems, different bugs), investigating them sequentially wastes time. Each investigation is independent and can happen in parallel.
8
9	Core principle: Dispatch one agent per independent problem domain. Let them work concurrently.
10
11	## When to Use
12
13	```dot
14	digraph when_to_use {
15	"Multiple failures?" [shape=diamond];
16	"Are they independent?" [shape=diamond];
17	"Single agent investigates all" [shape=box];
18	"One agent per problem domain" [shape=box];
19	"Can they work in parallel?" [shape=diamond];
20	"Sequential agents" [shape=box];
21	"Parallel dispatch" [shape=box];
22
23	"Multiple failures?" -> "Are they independent?" [label="yes"];
24	"Are they independent?" -> "Single agent investigates all" [label="no - related"];
25	"Are they independent?" -> "Can they work in parallel?" [label="yes"];
26	"Can they work in parallel?" -> "Parallel dispatch" [label="yes"];
27	"Can they work in parallel?" -> "Sequential agents" [label="no - shared state"];
28	}
29	```
30
31	Use when:
32	- 3+ test files failing with different root causes
33	- Multiple subsystems broken independently
34	- Each problem can be understood without context from others
35	- No shared state between investigations
36
37	Don't use when:
38	- Failures are related (fix one might fix others)
39	- Need to understand full system state
40	- Agents would interfere with each other
41
42	## The Pattern
43
44	### 1. Identify Independent Domains
45
46	Group failures by what's broken:
47	- File A tests: Tool approval flow
48	- File B tests: Batch completion behavior
49	- File C tests: Abort functionality
50
51	Each domain is independent - fixing tool approval doesn't affect abort tests.
52
53	### 2. Create Focused Agent Tasks
54
55	Each agent gets:
56	- Specific scope: One test file or subsystem
57	- Clear goal: Make these tests pass
58	- Constraints: Don't change other code
59	- Expected output: Summary of what you found and fixed
60
61	### 3. Dispatch in Parallel
62
63	Issue all three subagent dispatches in the same response — they run in parallel:
64
65	```text
66	Subagent (general-purpose): "Fix agent-tool-abort.test.ts failures"
67	Subagent (general-purpose): "Fix batch-completion-behavior.test.ts failures"
68	Subagent (general-purpose): "Fix tool-approval-race-conditions.test.ts failures"
69	# All three run concurrently.
70	```
71
72	Multiple dispatch calls in one response = parallel execution. One per response = sequential.
73
74	### 4. Review and Integrate
75
76	When agents return:
77	- Read each summary
78	- Verify fixes don't conflict
79	- Run full test suite
80	- Integrate all changes
81
82	## Agent Prompt Structure
83
84	Good agent prompts are:
85	1. Focused - One clear problem domain
86	2. Self-contained - All context needed to understand the problem
87	3. Specific about output - What should the agent return?
88
89	```markdown
90	Fix the 3 failing tests in src/agents/agent-tool-abort.test.ts:
91
92	1. "should abort tool with partial output capture" - expects 'interrupted at' in message
93	2. "should handle mixed completed and aborted tools" - fast tool aborted instead of completed
94	3. "should properly track pendingToolCount" - expects 3 results but gets 0
95
96	These are timing/race condition issues. Your task:
97
98	1. Read the test file and understand what each test verifies
99	2. Identify root cause - timing issues or actual bugs?
100	3. Fix by:
101	- Replacing arbitrary timeouts with event-based waiting
102	- Fixing bugs in abort implementation if found
103	- Adjusting test expectations if testing changed behavior
104
105	Do NOT just increase timeouts - find the real issue.
106
107	Return: Summary of what you found and what you fixed.
108	```
109
110	## Common Mistakes
111
112	❌ Too broad: "Fix all the tests" - agent gets lost
113	✅ Specific: "Fix agent-tool-abort.test.ts" - focused scope
114
115	❌ No context: "Fix the race condition" - agent doesn't know where
116	✅ Context: Paste the error messages and test names
117
118	❌ No constraints: Agent might refactor everything
119	✅ Constraints: "Do NOT change production code" or "Fix tests only"
120
121	❌ Vague output: "Fix it" - you don't know what changed
122	✅ Specific: "Return summary of root cause and changes"
123
124	## When NOT to Use
125
126	Related failures: Fixing one might fix others - investigate together first
127	Need full context: Understanding requires seeing entire system
128	Exploratory debugging: You don't know what's broken yet
129	Shared state: Agents would interfere (editing same files, using same resources)
130
131	## Real Example from Session
132
133	Scenario: 6 test failures across 3 files after major refactoring
134
135	Failures:
136	- agent-tool-abort.test.ts: 3 failures (timing issues)
137	- batch-completion-behavior.test.ts: 2 failures (tools not executing)
138	- tool-approval-race-conditions.test.ts: 1 failure (execution count = 0)
139
140	Decision: Independent domains - abort logic separate from batch completion separate from race conditions
141
142	Dispatch:
143	```
144	Agent 1 → Fix agent-tool-abort.test.ts
145	Agent 2 → Fix batch-completion-behavior.test.ts
146	Agent 3 → Fix tool-approval-race-conditions.test.ts
147	```
148
149	Results:
150	- Agent 1: Replaced timeouts with event-based waiting
151	- Agent 2: Fixed event structure bug (threadId in wrong place)
152	- Agent 3: Added wait for async tool execution to complete
153
154	Integration: All fixes independent, no conflicts, full suite green
155
156	## Verification
157
158	After agents return:
159	1. Review each summary - Understand what changed
160	2. Check for conflicts - Did agents edit same code?
161	3. Run full suite - Verify all fixes work together
162	4. Spot check - Agents can make systematic errors

6 picks

Type

Run `npx skills use "https://github.com/obra/superpowers" --skill "obra/superpowers/dispatching-parallel-agents"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.

Use the skills in "https://github.com/obra/superpowers" that are relevant to the current task. Run `npx skills add "https://github.com/obra/superpowers"` and select the relevant skills, then follow their instructions.

dispatching-parallel-agents

How to install dispatching-parallel-agents?

Related

dispatching-parallel-agents

How to install dispatching-parallel-agents?

Related