Generate a Claude Agent Skill from a local code directory. Creates a skill package containing SKILL.md (entry point with metadata) and references/ folder with summary.md, project-structure.md, files.md, and optionally tech-stacks.md. This tool creates Project Skills in <project>/.claude/skills/<n...
AI agents use generate_skill to create or update resources in Repomix — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Repomix environment.
The tool creates and writes new files and directories in a reversible manner (.claude/skills directory structure). While it modifies the local codebase, the changes are not destructive—files can be edited or deleted. This is a Write operation rather than Execute (does not run arbitrary code) or Destructive (changes are reversible).
From the tool's definition Tool 'generates' a skill package, 'creates' a skill in .claude/skills/<name>/ directory, and 'Creates a skill package containing SKILL.md and references/ folder'.
Risk signalsBulk/mass operation — affects multiple targets
Documented attack patterns abuse exactly the kind of access generate_skill gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Repomix, and nothing reaches the server without passing your rules. This is the rule we recommend for generate_skill:
{
"version": "1",
"default": "deny",
"tools": {
"generate_skill": {
"limits": [
{
"counter": "generate_skill_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}generate_skill stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Generate a Claude Agent Skill from a local code directory. Creates a skill package containing SKILL.md (entry point with metadata) and references/ folder with summary.md, project-structure.md, files.md, and optionally tech-stacks.md. This tool creates Project Skills in <project>/.claude/skills/<name>/, which are shared with the team via version control. Output Structure: .claude/skills/<skill-name>/ ├── SKILL.md # Entry point with usage guide └── references/ ├── summary.md # Purpose, format, and statistics ├── project-structure.md # Directory tree with line counts ├── files.md # All file contents └── tech-stacks.md # Languages, frameworks, dependencies (if detected) Example Path: /path/to/project/.claude/skills/repomix-reference-myproject/. It is categorised as a Write tool in the Repomix MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Repomix MCP server in PolicyLayer and add a rule for generate_skill: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Repomix. Nothing to install.
generate_skill is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the generate_skill rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for generate_skill. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
generate_skill is provided by the Repomix MCP server (yamadashy/repomix). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 8 Repomix tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
8 Repomix tools catalogued and risk-classified — across an index of 42,500+ MCP servers.