Menu
One imperative sentence on what this skill does + concrete trigger conditions. Be pushy — list contexts including ones where the user doesn't name the domain ("even when phrased as 'X' or 'Y'"). Pack relevant keywords. Stay under 1024 characters.
Create or edit AgentSync configuration — AGENTS.md, rules, skills, commands, subagents, settings, MCP servers, hooks, or per-tool configs. Use this skill when adding a rule, creating or scaffolding a skill, writing a slash command, defining a subagent persona, editing permissions, configuring an MCP server, setting up the `.ai/src/` directory, or running `agentsync sync` / `add` / `customize` / `resolve` / `simplify` / `profile` — even when the user does not name "AgentSync" explicitly but is editing files in `.ai/src/`, `.claude/`, `.cursor/`, or another tool-config directory.
Create a well-structured git commit for staged or unstaged changes — analyze the diff, match the project's existing commit style, write a clear message, and create the commit. Use this skill when the user asks to commit, save changes, check in code, write a commit message, prepare changes for push, or wrap up a piece of work — even when they don't say "commit" explicitly (e.g. "let's save this", "ship it", "сохрани изменения").
Design, debug, and refine prompts for LLM-based coding tools, agents, and pipelines. Use this skill when writing or rewriting a system prompt, agent persona, slash-command body, rule, or skill description; tuning a prompt that produces vague, incomplete, off-format, hallucinated, or over-/under-triggering output; choosing among techniques like XML tags, few-shot examples, role-setting, reasoning steps, or recap; debugging a model that ignores instructions, leaks information, or over-engineers solutions; or designing the persona, autonomy posture, or tool-use rules of an autonomous agent — even when the user does not explicitly mention "prompt engineering" or "system prompt". Not for AgentSync file structure, frontmatter, or scaffolding new rule/skill/command files — use the `agentsync` skill for that.
Design, debug, and refine prompts for LLM-based coding tools, agents, and pipelines. Use this skill when writing or rewriting a system prompt, agent persona, slash-command body, rule, or skill description; tuning a prompt that produces vague, incomplete, off-format, hallucinated, or over-/under-triggering output; choosing among techniques like XML tags, few-shot examples, role-setting, reasoning steps, or recap; debugging a model that ignores instructions, leaks information, or over-engineers solutions; or designing the persona, autonomy posture, or tool-use rules of an autonomous agent — even when the user does not explicitly mention "prompt engineering" or "system prompt". Not for AgentSync file structure, frontmatter, or scaffolding new rule/skill/command files — use the `agentsync` skill for that.
Investigate and fix a GitHub issue in this repo — pull the issue with `gh`, trace it to the root cause across `lib/` and `bin/`, fix portably, and cover it with a bats regression test. Use this skill when the user asks to fix, resolve, or tackle a GitHub issue (by number or URL), debug a reported bug from the tracker, or work through an issue end-to-end — including phrasings like "fix #42", "tackle issue 17", "resolve this GH issue", "почини issue #N", or "разберись с багом из тикета".
Create or edit AgentSync configuration — AGENTS.md, rules, skills, commands, subagents, settings, MCP servers, hooks, or per-tool configs. Use this skill when adding a rule, creating or scaffolding a skill, writing a slash command, defining a subagent persona, editing permissions, configuring an MCP server, setting up the `.ai/src/` directory, or running `agentsync sync` / `add` / `customize` / `resolve` / `simplify` — even when the user does not name "AgentSync" explicitly but is editing files in `.ai/src/`, `.claude/`, `.cursor/`, or another tool-config directory.
| name | content |
| description | One imperative sentence on what this skill does + concrete trigger conditions. Be pushy — list contexts including ones where the user doesn't name the domain ("even when phrased as 'X' or 'Y'"). Pack relevant keywords. Stay under 1024 characters. |
[One line: what this skill does and when the agent should invoke it.]
[Optional. Only include if you've created references/ files. Each line states a CONCRETE trigger condition, not a vague "see references/".]
references/X.md — read when [concrete trigger condition][The single highest-leverage section. Concrete corrections to wrong assumptions, not generic advice.]