Skip to main content
Glama

agy_spawn_swarm

Read-onlyIdempotent

Dispatch multiple read-only Antigravity research subagents across non-overlapping directory partitions to explore independent subsystems in one turn, then aggregate their markdown reports.

Instructions

Dispatch multiple read-only Antigravity subagents using gemini-3.8-flash-low or gemini-3.8-flash-medium across strictly disjoint directory partitions. SERIALIZATION & ASYNC BUDGET: Tasks run sequentially one after another (one at a time, AGY_MAX_CONCURRENCY = 1) and share the same ~8-minute soft budget ceiling; keep each task prompt bounded or fetch any in-flight task result by job_id via agy_job_result. WHEN TO USE: Exploring multiple independent subsystems or directories in a single turn. DO NOT USE FOR: Single-directory research (use agy_delegate_research), single-file signature extraction (use local_file_scout), or overlapping directory paths (rejected by partition validator). RETURNS: Aggregated markdown report per subagent task (or async job_id handles for tasks still running at soft budget) plus telemetry footer.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "tasks": "<tasks>"
    },
    "name": "agy_spawn_swarm"
  }
}

Input Schema

NameRequiredDescriptionDefault
cwdNoOptional workspace root directory path.
tasksYesArray of subagent tasks with strictly non-overlapping directory globs (dispatched sequentially one after another).
soft_timeout_msNoOptional soft budget in milliseconds before returning async job_id handles for any remaining tasks (default 540000 = 540s).

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changedv1.3.0
    • addedInput schema / properties / soft_timeout_ms
      Added value: +{
      +  "description": "Optional soft budget in milliseconds before returning async job_id handles for any remaining tasks (default 540000 = 540s).",
      +  "maximum": 590000,
      +  "minimum": 100,
      +  "type": "number"
      +}
    • changedInput schema / properties / tasks / description
      Previous value: -"Array of subagent tasks with strictly non-overlapping directory globs."New value: +"Array of subagent tasks with strictly non-overlapping directory globs (dispatched sequentially one after another)."
  2. Changed12 schema fields changedv1.2.1
    • addedInput schema / additionalProperties
      Added value: +false
    • changedInput schema / properties / cwd / description
      Previous value: -"Optional workspace root directory."New value: +"Optional workspace root directory path."
    • changedInput schema / properties / tasks / description
      Previous value: -"Array of parallel subagent tasks with disjoint path partitions."New value: +"Array of subagent tasks with strictly non-overlapping directory globs."
    • addedInput schema / properties / tasks / items / additionalProperties
      Added value: +false
    • changedInput schema / properties / tasks / items / properties / id / description
      Previous value: -"Unique task identifier."New value: +"Unique task identifier (e.g. \"auth-worker\")."
    • addedInput schema / properties / tasks / items / properties / id / minLength
      Added value: +1
    • addedInput schema / properties / tasks / items / properties / paths / minItems
      Added value: +1
    • changedInput schema / properties / tasks / items / properties / prompt / description
      Previous value: -"Task instructions for this subagent."New value: +"Read-only research instructions for this subagent partition."
    • addedInput schema / properties / tasks / items / properties / prompt / minLength
      Added value: +1
    • addedInput schema / properties / tasks / items / properties / tier / default
      Added value: +"flash-low"
    • changedInput schema / properties / tasks / items / properties / tier / description
      Previous value: -"Swarm tier (defaults to flash-low)."New value: +"Swarm model tier (defaults to flash-low)."
    • addedInput schema / properties / tasks / minItems
      Added value: +1
  3. First observedv0.1.0

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes well past the annotations (readOnly/idempotent/non-destructive) by disclosing the concurrency model (AGY_MAX_CONCURRENCY = 1, tasks run one at a time), the shared ~8-minute soft budget, the async job_id escape hatch for in-flight tasks, and the fact that overlapping paths are rejected by a partition validator. The RETURNS clause also sets expectations for mixed inline-report and job-handle responses.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Labeled sections (SERIALIZATION & ASYNC BUDGET, WHEN TO USE, DO NOT USE FOR, RETURNS) front-load the core action and make scanning fast. Slight redundancy in restating sequential execution three ways ('sequentially one after another (one at a time, AGY_MAX_CONCURRENCY = 1)'), but the emphasis is defensible given the cost of misreading it.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a fan-out tool with no output schema, the description covers the model choice, the disjointness invariant, timeout behavior, the async retrieval path via agy_job_result, and the aggregated-markdown return shape. An agent has everything needed to compose valid tasks and interpret results.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already documents cwd, tasks, and soft_timeout_ms; the description adds cross-referencing value by mapping the tier enum to concrete models (gemini-3.8-flash-low/medium) and by tying the ~8-minute soft budget to the async job_id fallback. It stops short of adding per-parameter syntax beyond what the schema states, so it is an add-on rather than a substitute.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource (dispatch multiple read-only subagents) plus the defining constraint that makes this tool distinct: strictly disjoint directory partitions and sequential execution. An agent can distinguish it from agy_delegate_research or local_file_scout without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit WHEN TO USE (exploring multiple independent subsystems/directories in one turn) and DO NOT USE FOR clauses that name concrete alternatives (agy_delegate_research for single-directory, local_file_scout for single-file signatures) plus the invalid case of overlapping paths. Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.