antigravity-swarm

Deploys autonomous sub-agents to perform tasks in the Antigravity IDE. Supports both manual dispatch and dynamic "Auto-Hiring" of agent teams.

Quality

41%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Critical

Do not install without reviewing

Optimize this skill with Tessl

npx tessl skill review --optimize ./public/skills/0xnagato/antigravity-swarm/SKILL.md

Quality

Discovery

32%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description identifies a specific platform (Antigravity IDE) and two operational modes but fails to specify what concrete tasks the sub-agents perform and lacks any explicit trigger guidance for when Claude should select this skill. The vague 'perform tasks' phrasing significantly weakens both specificity and completeness.

Suggestions

Add an explicit 'Use when...' clause with trigger terms like 'deploy agents', 'hire agents', 'delegate tasks', 'run sub-agents', 'Antigravity', 'agent team'.

Replace 'perform tasks' with specific concrete actions the sub-agents can do, e.g., 'run code, execute tests, refactor files, or handle multi-step workflows'.

Include natural keyword variations users might say, such as 'spawn workers', 'parallel execution', 'delegate work', or 'multi-agent'.

Dimension	Reasoning	Score
Specificity	Names the domain (Antigravity IDE, sub-agents) and mentions two modes ('manual dispatch' and 'Auto-Hiring'), but doesn't list concrete actions beyond 'perform tasks', which is vague about what those tasks actually are.	2 / 3
Completeness	Describes what it does at a high level but has no explicit 'Use when...' clause or equivalent trigger guidance, and the 'what' is itself vague ('perform tasks'). The missing 'when' clause caps this at 2 per the rubric, and the weak 'what' brings it to 1.	1 / 3
Trigger Term Quality	Includes some relevant terms like 'sub-agents', 'Auto-Hiring', 'Antigravity IDE', and 'deploy', but misses common user-facing variations like 'agent', 'delegate', 'spawn', 'worker', or 'parallel tasks' that users might naturally say.	2 / 3
Distinctiveness Conflict Risk	The mention of 'Antigravity IDE' and 'Auto-Hiring' provides some distinctiveness, but 'deploys autonomous sub-agents to perform tasks' is broad enough to potentially overlap with other agent orchestration or task delegation skills.	2 / 3
	Total	7 / 12 Passed

Implementation

50%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The skill provides a reasonable overview of a sub-agent dispatch system with concrete examples and clear section organization. However, it suffers from unnecessary philosophical explanations that inflate token usage, incomplete actionability in the monitoring workflow, and missing validation/error-recovery steps in multi-step processes. The tool interface descriptions are somewhat ambiguous (MCP tool vs CLI command).

Suggestions

Remove or move the FAQ & Philosophy section to a separate PHILOSOPHY.md file—Claude doesn't need to be convinced why sub-agents are useful, it needs to know how to use them.

Add explicit error handling and validation steps to the workflows: what to check after dispatch, how to detect and retry failed agents, and what constitutes a successful completion.

Clarify the tool interface: are `dispatch_subagent` and `run_mission` MCP tools, CLI commands, or Python functions? The examples use `run_command` wrapping Python scripts but the tool descriptions read like MCP tool definitions.

Add executable code for the monitoring/polling step in IDE Agent mode instead of the vague 'Poll the JSON log files' instruction.

Dimension	Reasoning	Score
Conciseness	The skill includes some unnecessary explanations (e.g., the FAQ section explaining why sub-agents are useful, the 'Is this truly parallel?' section, and the philosophy discussion). These are concepts Claude can reason about without being told. The core tool documentation is reasonably efficient, but the FAQ adds ~30% bloat.	2 / 3
Actionability	Provides concrete commands and examples for dispatching agents and running missions, but key details are incomplete—e.g., the `dispatch_subagent` and `run_mission` tools are described but it's unclear if these are MCP tools, Python functions, or CLI commands (the examples use `run_command` wrapping Python scripts, which differs from the tool descriptions). The IDE Agent mode gives a workflow but the monitoring/polling step lacks executable code.	2 / 3
Workflow Clarity	The IDE Agent mode outlines a spawn→monitor→visualize sequence, and Mission Mode shows a two-step generate→run flow. However, there are no validation checkpoints or error recovery steps. What happens if a sub-agent fails? The FAQ mentions fault tolerance but the workflow doesn't include retry logic or validation steps. The warning about not modifying files is good but reactive rather than procedural.	2 / 3
Progressive Disclosure	The content is structured with clear sections (Tools, Usage Modes, Examples, FAQ), which is good. However, the FAQ/Philosophy section is inlined when it could be in a separate file, and there are no references to external documentation for the Manus Protocol, planner.py internals, or advanced configuration. No bundle files are provided to support the referenced scripts (planner.py, orchestrator.py, dispatch_agent.py), making it hard to verify path accuracy.	2 / 3
	Total	8 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: Demerzels-lab/elsamultiskillagent
Commit: f45fcb5

Reviewed: 23 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.