Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/cboone/agent-harness-plugins/create-pluginnpx skills add cboone/agent-harness-plugins --skill create-plugingit clone --depth 1 https://github.com/cboone/agent-harness-pluginsWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00081 | $0.02579 |
| Opus 5 | $0.00041 | $0.01290 |
| Sonnet 5 | $0.00016 | $0.00516 |
| Haiku 4.5 | $0.00008 | $0.00258 |
Grade A, and why
create-plugin scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 212 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create Plugin
Create a new plugin for this repository following established conventions.
Workflow
1. Determine Plugin Type
Infer the plugin type from the user's request:
- Skills plugin: Provides instructions and workflows that Claude Code follows (e.g., style guides, multi-step procedures). Most plugins are this type.
- Command plugin: Provides slash commands that users invoke explicitly (e.g.,
/example-command). Commands are structured Markdown files with frontmatter, a workflow, and optional reference templates. - Hooks plugin: Provides event-driven shell commands that run automatically in response to Claude Code lifecycle events (e.g., notifications on task completion).
- Combinations: A plugin can provide any combination of skills, commands, and hooks.
Inference heuristic: "create a command", "add a slash command", "add a /something command" implies a command plugin. "create a skill", "add a style guide", "add a workflow" implies a skills plugin. "create a hook", "add a notification" implies a hooks plugin.
If the request doesn't imply a type (e.g., just "create a plugin"), ask. If ambiguous, default to a skills plugin.
2. Choose a Name
The plugin name must be:
- Kebab-case (e.g.,
write-go-code,suggest-next-issue) - Verb-noun preferred (e.g.,
create-worktree-from-issue,resolve-copilot-pr-feedback) - Descriptive of what the plugin does
- Unique within the
plugins/directory
If the user provided a name, use it. Otherwise, generate a descriptive name from the plugin's purpose and proceed.
3. Create Directory Structure
No manual mkdir is needed — the Write tool creates parent directories automatically when writing files. The directories below are created implicitly when their first file is written in the subsequent steps.
Skills Plugin
plugins/PLUGIN-NAME/.claude-plugin/
plugins/PLUGIN-NAME/skills/PLUGIN-NAME/
Add a references/ subdirectory if the skill needs supplementary documentation:
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 212 lines · 81 tokens per session scan A fb2764d3d7ad
create-plugin is a skill published in the GitHub repository cboone/agent-harness-plugins (2 stars, last pushed 1mo ago), licensed MIT. It adds 81 tokens to every session and 2,579 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
create-skill
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
subagent
Spawn and manage autonomous background agents for parallel work.
skills
Manage and create Wolffish's own skills (capabilities) — list, search, enable, disable, delete, and author brand-new skills at runtime.
viral-short-form
Brainstorm and write high-retention short-form video and carousel content for TikTok, Reels, and YouTube Shorts. Use whenever someone wants viral hook ideas, a video script or outline, content concepts for a product or topic, or wants to critique and improve a draft hook or script. Works for any short-form format…
decision-review
Private evidence-grounded decision-review method read directly by the mainframe decision reviewer. Not a primary-session capability.
skill-creator
Scaffold a new Claude Code skill with SKILL.md, frontmatter, and bundled resources. Use when creating a custom skill, standardizing skill structure across a team, or packaging a skill for distribution.