Files

spaceman1412 fd4c3c2fc7 Add new skills and utilities for enhanced writing and testing

- Introduced graphviz conventions for visualizing process flows in writing skills.
- Added a comprehensive guide on persuasion principles to improve skill design effectiveness.
- Implemented a script to render graphviz diagrams from markdown files to SVG format.
- Created a detailed reference for testing skills with subagents, emphasizing TDD principles.
- Established a task tracker template for live task management.
- Developed a shell script to check the integrity of the antigravity profile and required files.
- Added test scripts to validate the initialization of agent projects.
- Created workflows for brainstorming, executing plans, and writing plans to streamline processes.

2026-02-26 15:16:32 +07:00

2.3 KiB

Raw Blame History

Superpowers for Antigravity

You have superpowers.

This profile adapts Superpowers workflows for Antigravity with strict single-flow execution.

Core Rules

Prefer local skills in .agent/skills/<skill-name>/SKILL.md.
Execute one core task at a time with task_boundary.
Use browser_subagent only for browser automation tasks.
Track checklist progress in <project-root>/docs/plans/task.md (table-only live tracker).
Keep changes scoped to the requested task and verify before completion claims.

Tool Translation Contract

When source skills reference legacy tool names, use these Antigravity equivalents:

Legacy assistant/platform names -> Antigravity
Task tool -> browser_subagent for browser tasks, otherwise sequential task_boundary
Skill tool -> view_file ~/.gemini/skills/<skill-name>/SKILL.md (or project-local .agent/skills/<skill-name>/SKILL.md)
TodoWrite -> update <project-root>/docs/plans/task.md task list
File operations -> view_file, write_to_file, replace_file_content, multi_replace_file_content
Directory listing -> list_dir
Code structure -> view_file_outline, view_code_item
Search -> grep_search, find_by_name
Shell -> run_command
Web fetch -> read_url_content
Web search -> search_web
Image generation -> generate_image
User communication during tasks -> notify_user
MCP tools -> mcp_* tool family

Skill Loading

First preference: project skills at .agent/skills.
Second preference: user skills at ~/.gemini/skills.
If both exist, project-local skills win for this profile.
Optional parity assets may exist at .agent/workflows/* and .agent/agents/* as entrypoint shims/reference profiles.
These assets do not change the strict single-flow execution requirements in this file.

Single-Flow Execution Model

Do not dispatch multiple coding agents in parallel.
Decompose large work into ordered, explicit steps.
Keep exactly one active task at a time in <project-root>/docs/plans/task.md.
If browser work is required, isolate it in a dedicated browser step.

Verification Discipline

Before saying a task is done:

Run the relevant verification command(s).
Confirm exit status and key output.
Update <project-root>/docs/plans/task.md.
Report evidence, then claim completion.

2.3 KiB Raw Blame History