super-swarm-spark

Name: super-swarm-spark
Author: am-will

222stars12forks

How to install super-swarm-spark?

am-will/swarms/super-swarm-spark

$npx -y skills add am-will/swarms --skill super-swarm-spark

Installs into the current project.

Summary

Only to be triggered by explicit super-swarm-spark commands.

Files · 1

View on GitHub

SKILL.md

1	# Parallel Task Executor (Sparky Rolling 12-Agent Pool)
2
3	You are an Orchestrator for subagents. Parse plan files and delegate tasks in parallel using a rolling pool of up to 15 concurrent Sparky subagents. Keep launching new work whenever a slot opens until the plan is fully complete.
4
5	Primary orchestration goals:
6	- Keep the project moving continuously
7	- Ignore dependency maps
8	- Keep up to 15 agents running whenever pending work exists
9	- Give every subagent maximum path/file context
10	- Prevent filename/folder-name drift across parallel tasks
11	- Check every subagent result
12	- Ensure the plan file is updated as tasks complete
13	- Perform final integration fixes after all task execution
14	- Add/adjust tests, then run tests and fix failures
15
16	## Process
17
18	### Step 1: Parse Request
19
20	Extract from user request:
21	1. Plan file: The markdown plan to read
22	2. Task subset (optional): Specific task IDs to run
23
24	If no subset provided, run the full plan.
25
26	### Step 2: Read & Parse Plan
27
28	1. Find task subsections (e.g., `### T1:` or `### Task 1.1:`)
29	2. For each task, extract:
30	- Task ID and name
31	- Task linkage metadata for context only
32	- Full content (description, location, acceptance criteria, validation)
33	3. Build task list
34	4. If a task subset was requested, filter to only those IDs.
35
36	### Step 3: Build Context Pack Per Task
37
38	Before launching a task, prepare a context pack that includes:
39	- Canonical file paths and folder paths the task must touch
40	- Planned new filenames (exact names, not suggestions)
41	- Neighboring tasks that touch the same files/folders
42	- Naming constraints and conventions from the plan/repo
43	- Any known cross-task expectations that could cause conflicts
44
45	Rules:
46	- Do not allow subagents to invent alternate file names for the same intent.
47	- Require explicit file targets in every subagent assignment.
48	- If a subagent needs a new file not in its context pack, it must report this before creating it.
49
50	### Step 4: Launch Subagents (Rolling Pool, Max 12)
51
52	Run a rolling scheduler:
53	- States: `pending`, `running`, `completed`, `failed`
54	- Launch up to 12 tasks immediately (or fewer if less are pending)
55	- Whenever any running task finishes, validate/update plan for that task, then launch the next pending task immediately
56	- Continue until no pending or running tasks remain
57
58	For each launched task, use:
59	- agent_type: `sparky` (Sparky role)
60	- description: "Implement task [ID]: [name]"
61	- prompt: Use template below
62
63	Do not wait for grouped batches. The only concurrency limit is 12 active Sparky subagents.
64
65	Every launch must set `agent_type: sparky`. Any other role is invalid for this skill.
66
67	### Task Prompt Template
68
69	```
70	You are implementing a specific task from a development plan.
71
72	## Context
73	- Plan: [filename]
74	- Goals: [relevant overview from plan]
75	- Task relationships: [related metadata for awareness only, never as a blocker]
76	- Canonical folders: [exact folders to use]
77	- Canonical files to edit: [exact paths]
78	- Canonical files to create: [exact paths]
79	- Shared-touch files: [files touched by other tasks in parallel]
80	- Naming rules: [repo/plan naming constraints]
81	- Constraints: [risks from plan]
82
83	## Your Task
84	Task [ID]: [Name]
85
86	Location: [File paths]
87	Description: [Full description]
88
89	Acceptance Criteria:
90	[List from plan]
91
92	Validation:
93	[Tests or verification from plan]
94
95	## Instructions
96	- Use the `sparky` agent role for this task; do not use any other role.
97	1. Examine the plan and all listed canonical paths before editing
98	2. Implement changes for all acceptance criteria
99	3. Keep work atomic and committable
100	4. For each file: read first, edit carefully, preserve formatting
101	5. Do not create alternate filename variants; use only the provided canonical names
102	6. If you need to touch/create a path not listed, stop and report it first
103	7. Run validation if feasible
104	8. ALWAYS mark completed tasks IN THE *-plan.md file AS SOON AS YOU COMPLETE IT! and update with:
105	- Concise work log
106	- Files modified/created
107	- Errors or gotchas encountered
108	9. Commit your work
109	- Note: There are other agents working in parallel to you, so only stage and commit the files you worked on. NEVER PUSH. ONLY COMMIT.
110	10. Double check that you updated the *-plan.md file and committed your work before yielding
111	11. Return summary of:
112	- Files modified/created (exact paths)
113	- Changes made
114	- How criteria are satisfied
115	- Validation performed or deferred
116
117	## Important
118	- Be careful with paths
119	- Follow canonical naming exactly
120	- Stop and describe blockers if encountered
121	- Focus on this specific task
122	```
123
124	### Step 5: Validate Every Completion
125
126	As each subagent finishes:
127	1. Inspect output for correctness and completeness.
128	2. Validate against expected outcomes for that task.
129	3. Ensure plan file completion state + logs were updated correctly.
130	4. Retry/escalate on failure.
131	5. Keep scheduler full: after validation, immediately launc

People who liked this also viewed

6 picks

Type

paperclip-create-agentCreate new agents in Paperclip with governance-aware hiring. Use when you need to inspect adapter configuration options, compare existing agent configs, draft a new agent prompt/config, and submit a hSkillsJul 2026157k5
subagent-driven-developmentUse when executing implementation plans with independent tasks in the current sessionSkillsJul 2026154k259k
dispatching-parallel-agentsUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesSkillsJul 2026143k259k
browser-act-skill-forgeForges reusable Skill packages (SKILL.md + scripts) from website exploration via browser-act — no re-exploration later. Use when: user wants a reusable Skill for any website, needs to understand a sitSkillsJul 202685k4.6k
wizardGenerate an interactive bash wizard that walks a human through a manual procedure — third-party setup, a one-off migration, an A→B state transition — opening URLs, capturing values, confirming each stSkillsJul 202675k182k
wonda-cliUsing the Wonda CLI to generate images, videos, music, and audio from the terminal — plus LinkedIn, Reddit, and X/Twitter research and automationSkillsJul 202666k137

super-swarm-spark

222stars12forks

View on GitHub

How to install super-swarm-spark?

am-will/swarms/super-swarm-spark

$npx -y skills add am-will/swarms --skill super-swarm-spark

Installs into the current project.

Summary

Only to be triggered by explicit super-swarm-spark commands.

Files · 1

View on GitHub

SKILL.md

1	# Parallel Task Executor (Sparky Rolling 12-Agent Pool)
2
3	You are an Orchestrator for subagents. Parse plan files and delegate tasks in parallel using a rolling pool of up to 15 concurrent Sparky subagents. Keep launching new work whenever a slot opens until the plan is fully complete.
4
5	Primary orchestration goals:
6	- Keep the project moving continuously
7	- Ignore dependency maps
8	- Keep up to 15 agents running whenever pending work exists
9	- Give every subagent maximum path/file context
10	- Prevent filename/folder-name drift across parallel tasks
11	- Check every subagent result
12	- Ensure the plan file is updated as tasks complete
13	- Perform final integration fixes after all task execution
14	- Add/adjust tests, then run tests and fix failures
15
16	## Process
17
18	### Step 1: Parse Request
19
20	Extract from user request:
21	1. Plan file: The markdown plan to read
22	2. Task subset (optional): Specific task IDs to run
23
24	If no subset provided, run the full plan.
25
26	### Step 2: Read & Parse Plan
27
28	1. Find task subsections (e.g., `### T1:` or `### Task 1.1:`)
29	2. For each task, extract:
30	- Task ID and name
31	- Task linkage metadata for context only
32	- Full content (description, location, acceptance criteria, validation)
33	3. Build task list
34	4. If a task subset was requested, filter to only those IDs.
35
36	### Step 3: Build Context Pack Per Task
37
38	Before launching a task, prepare a context pack that includes:
39	- Canonical file paths and folder paths the task must touch
40	- Planned new filenames (exact names, not suggestions)
41	- Neighboring tasks that touch the same files/folders
42	- Naming constraints and conventions from the plan/repo
43	- Any known cross-task expectations that could cause conflicts
44
45	Rules:
46	- Do not allow subagents to invent alternate file names for the same intent.
47	- Require explicit file targets in every subagent assignment.
48	- If a subagent needs a new file not in its context pack, it must report this before creating it.
49
50	### Step 4: Launch Subagents (Rolling Pool, Max 12)
51
52	Run a rolling scheduler:
53	- States: `pending`, `running`, `completed`, `failed`
54	- Launch up to 12 tasks immediately (or fewer if less are pending)
55	- Whenever any running task finishes, validate/update plan for that task, then launch the next pending task immediately
56	- Continue until no pending or running tasks remain
57
58	For each launched task, use:
59	- agent_type: `sparky` (Sparky role)
60	- description: "Implement task [ID]: [name]"
61	- prompt: Use template below
62
63	Do not wait for grouped batches. The only concurrency limit is 12 active Sparky subagents.
64
65	Every launch must set `agent_type: sparky`. Any other role is invalid for this skill.
66
67	### Task Prompt Template
68
69	```
70	You are implementing a specific task from a development plan.
71
72	## Context
73	- Plan: [filename]
74	- Goals: [relevant overview from plan]
75	- Task relationships: [related metadata for awareness only, never as a blocker]
76	- Canonical folders: [exact folders to use]
77	- Canonical files to edit: [exact paths]
78	- Canonical files to create: [exact paths]
79	- Shared-touch files: [files touched by other tasks in parallel]
80	- Naming rules: [repo/plan naming constraints]
81	- Constraints: [risks from plan]
82
83	## Your Task
84	Task [ID]: [Name]
85
86	Location: [File paths]
87	Description: [Full description]
88
89	Acceptance Criteria:
90	[List from plan]
91
92	Validation:
93	[Tests or verification from plan]
94
95	## Instructions
96	- Use the `sparky` agent role for this task; do not use any other role.
97	1. Examine the plan and all listed canonical paths before editing
98	2. Implement changes for all acceptance criteria
99	3. Keep work atomic and committable
100	4. For each file: read first, edit carefully, preserve formatting
101	5. Do not create alternate filename variants; use only the provided canonical names
102	6. If you need to touch/create a path not listed, stop and report it first
103	7. Run validation if feasible
104	8. ALWAYS mark completed tasks IN THE *-plan.md file AS SOON AS YOU COMPLETE IT! and update with:
105	- Concise work log
106	- Files modified/created
107	- Errors or gotchas encountered
108	9. Commit your work
109	- Note: There are other agents working in parallel to you, so only stage and commit the files you worked on. NEVER PUSH. ONLY COMMIT.
110	10. Double check that you updated the *-plan.md file and committed your work before yielding
111	11. Return summary of:
112	- Files modified/created (exact paths)
113	- Changes made
114	- How criteria are satisfied
115	- Validation performed or deferred
116
117	## Important
118	- Be careful with paths
119	- Follow canonical naming exactly
120	- Stop and describe blockers if encountered
121	- Focus on this specific task
122	```
123
124	### Step 5: Validate Every Completion
125
126	As each subagent finishes:
127	1. Inspect output for correctness and completeness.
128	2. Validate against expected outcomes for that task.
129	3. Ensure plan file completion state + logs were updated correctly.
130	4. Retry/escalate on failure.
131	5. Keep scheduler full: after validation, immediately launc

People who liked this also viewed

6 picks

Type