codely
Codely’s plugins to level up your code.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Check it deeper on your machine
It has changed since its scan on 11 Nov 2026.
As a Claude Code plugin, install it and then ask Claude Code:
/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/CodelyTV/agent-harness with scanskillsafety
Or from a terminal, clone it and run it:
git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/CodelyTV/agent-harness'
It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.
Badge for your README
Markdown
[](/listings/codelytv--agent-harness)HTML
<a href="/listings/codelytv--agent-harness"><img src="/api/listings/codelytv--agent-harness/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
1. Install the skills with the `skills.sh` installer:
npx skills@latest add codelytv/agent-harness
2. Pick the skills you want and the coding agents you want to install them on.
3. Run a skill in your agent, for example `/codely-doc-create`.
4. Done. Your agent now follows Codely's conventions.
Alternative: Claude Code plugin
1. Add this repository as a plugin marketplace and install the plugin:
/plugin marketplace add CodelyTV/agent-harness /plugin install codely-skills@codely
2. Run a skill in Claude Code, for example `/codely-doc-create`.
3. Done. Every skill in this repo is now available in Claude Code.
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 48 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- {
- "name": "codely",
- "description": "Codely’s plugins to level up your code.",
- "owner": {
- "name": "Codely",
- "email": "support@codely.com",
- "url": "https://codely.com"
- },
- "plugins": [
- {
- "name": "codely-skills",
- "source": "./",
- "description": "Codely's team skills: planning, docs, OpenSpec, and git conventions.",
- "category": "engineering"
- },
- {
- "name": "ste100",
- "description": "Makes the agent answer in ASD-STE100 Simplified Technical English.",
- "version": "1.0.0",
- "source": "./.claude/plugins/ste100",
- "category": "productivity"
- },
- {
- "name": "ts7-lsp",
- "description": "TypeScript/JavaScript code intelligence via the TypeScript 7 native compiler LSP, much faster than tsserver on large monorepos",
- "version": "2.0.0",
- "source": "./.claude/plugins/ts7-lsp",
- "category": "development",
- "strict": false,
- "lspServers": {
- "ts7": {
- "command": "tsc",
- "args": ["--lsp", "-stdio"],
- "extensionToLanguage": {
- ".ts": "typescript",
- ".tsx": "typescriptreact",
- ".js": "javascript",
- ".jsx": "javascriptreact",
- ".mts": "typescript",
- ".cts": "typescript",
- ".mjs": "javascript",
- ".cjs": "javascript"
- }
- }
- }
- }
- ]
- }
README.md · README · 212 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- <p align="center">
- <a href="https://codely.com">
- <picture>
- <source media="(prefers-color-scheme: dark)" srcset="https://codely.com/logo/codely_logo-dark.svg">
- <source media="(prefers-color-scheme: light)" srcset="https://codely.com/logo/codely_logo-light.svg">
- <img alt="Codely logo" src="https://codely.com/logo/codely_logo.svg">
- </picture>
- </a>
- </p>
- <h1 align="center">🤖 Agent Harness</h1>
- <p align="center">
- <a href="https://github.com/CodelyTV"><img alt="Codely OS" src="https://img.shields.io/badge/Codely-OS-green?style=flat-square"></a>
- <a href="https://codely.com/cafe"><img alt="Codely Cafe News" src="https://img.shields.io/badge/Codely-Cafe_News-%235A3E2B?style=flat-square"></a>
- <a href="https://codely.com/newsletter"><img alt="Codely Newsletter" src="https://img.shields.io/badge/Codely-Newsletter-%23D3D3D3?style=flat-square"></a>
- <a href="https://youtube.com/codelytv"><img alt="Codely YouTube" src="https://img.shields.io/badge/Codely-YouTube-red?style=flat-square"></a>
- <a href="https://codely.com/courses"><img alt="Codely Courses" src="https://img.shields.io/badge/Codely-Courses-black?style=flat-square"></a>
- <a href="https://github.com/CodelyTV/agent-harness/stargazers"><img alt="GitHub Repo stars" src="https://img.shields.io/github/stars/codelytv/agent-harness"></a>
- </p>
- <p align="center">
- Skills, plugins and hooks so your agent writes the code you expect, the way you expect.<br>
- </p>
- Your agent writes code faster than you can review it. Instead of reading every line, turn your conventions into rules
- that fail in seconds: the agent corrects itself, and you review what the change does, not how it is written. Review outputs, not code.
- ## Workflow
- 1. **Make your docs deterministic** with `/deterministify docs`.
- 2. **Enforce your code conventions** with `/deterministify conventions`.
- 3. **Add a new rule** with `/create-rule` every time you correct the agent.
- 4. **Commit** with `/git-commit`.
- 5. **Open the pull request** with `/create-pull-request`, with before/after evidence of what the change does: screenshots, videos and architecture diagrams.
- For larger tasks, plan the work with `/create-plan` and implement it one phase at a time with
- `/implement-plan-phase`. Already using [OpenSpec](https://openspec.dev)? Run `/openspec-create-spec-from-code` first.
- ## Why
- ### Your docs are suggestions
- You wrote your conventions down, but the agent only follows them if it loads the right doc and reads it the way your
- team does.
- **The fix:** `/deterministify docs` turns every convention that code can check into a linter rule or a test.
- ### Your code does the same job in two ways
- A legacy way and a current way live together, and the agent copies whichever it reads first.
- **The fix:** `/deterministify conventions` finds them in the code and its history, and enforces the current way with a
- rule, after you confirm it.
- ### You correct the same thing again and again
- You fix the agent in one session, and it makes the same mistake in the next one.
- **The fix:** run `/create-rule` after the correction. It becomes a rule when a rule can check it, and a minimal doc
- with the reason when it cannot.
- ### The history is a mess
- Commit messages follow whatever the agent felt like that day.
- **The fix:** `/git-commit` writes Conventional Commits with the co-author trailer of the agent that made the change.
- ## Install
- ```bash
- npx skills@latest add codelytv/agent-harness
- ```
- Pick the skills you want and the agents to install them on. Then run one, for example `/create-rule`.
- <details>
- <summary>Claude Code plugin</summary>
- ```
- /plugin marketplace add CodelyTV/agent-harness
- /plugin install codely-skills@codely
- ```
- </details>
- ## Skills
- | Skill | Use it when | Variants |
- |--------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------|------------------------------------|
- | [`deterministify`](.agents/skills/deterministify/SKILL.md) | You want the conventions your repository already has enforced by linter rules and tests. | aliases |
- | [`create-rule`](.agents/skills/create-rule/SKILL.md) | You corrected the agent, or want a new convention enforced, and want a rule instead of a doc. | |
- | [`git-commit`](.agents/skills/git-commit/SKILL.md) | You want to commit your changes with a Conventional Commit message. | |
- | [`create-pull-request`](.agents/skills/create-pull-request/SKILL.md) | You open or update a pull request and want its description to follow a template with evidence. | |
- | [`create-plan`](.agents/skills/create-plan/SKILL.md) | You start a larger task and want a plan split into reviewable phases. | `-subagents`, `-github`, `-gitlab` |
- | [`implement-plan-phase`](.agents/skills/implement-plan-phase/SKILL.md) | You have a plan and want to implement its next phase. | `-github`, `-gitlab` |
- | [`openspec-create-spec-from-code`](.agents/skills/openspec-create-spec-from-code/SKILL.md) | You need the OpenSpec specs of behaviour that already exists in the code. | |
- | [`create-doc`](.agents/skills/create-doc/SKILL.md) | Deprecated: use `create-rule`. | |
- <details>
- <summary>Variants</summary>
- - aliases: [`docs-to-rules`](.agents/skills/docs-to-rules/SKILL.md) runs `deterministify docs`, and
- [`conventions-to-rules`](.agents/skills/conventions-to-rules/SKILL.md) runs `deterministify conventions`. Both
- need `deterministify` installed next to them.
- - `-subagents`: [`create-plan-subagents`](.agents/skills/create-plan-subagents/SKILL.md) delegates the codebase
- exploration to cheaper subagents running in parallel.
- - `-github`: [`create-plan-github`](.agents/skills/create-plan-github/SKILL.md) stores the plan as a parent issue with
- one sub-issue per phase, and [`implement-plan-phase-github`](.agents/skills/implement-plan-phase-github/SKILL.md)
- opens one pull request per phase.
- - `-gitlab`: [`create-plan-gitlab`](.agents/skills/create-plan-gitlab/SKILL.md) and
- [`implement-plan-phase-gitlab`](.agents/skills/implement-plan-phase-gitlab/SKILL.md) do the same with GitLab issues
- and merge requests.
- </details>
- <details>
- <summary><code>codely-spec-plan</code> OpenSpec schema</summary>
- An [OpenSpec custom schema](https://github.com/Fission-AI/OpenSpec/blob/main/docs/customization.md) that brings the
- `create-plan-subagents` flow to OpenSpec. Each change produces only two things to review:
- - **Lean specs**: one behaviour with SHALL per requirement and a single-line scenario.
- - **`plan.md`**: goal, decisions and one vertical slice phase per part a user can complete a goal with, each with its
- contracts and a test strategy that proves it with few tests. Its checkboxes are the task list that
- `/openspec-apply-change` works through one phase per run, committing each phase.
- Install it in your project and select it in `openspec/config.yaml`:
- ```bash
- npx degit CodelyTV/agent-harness/openspec/schemas/codely-spec-plan openspec/schemas/codely-spec-plan
- ```
- ```yaml
- schema: codely-spec-plan
- ```
- </details>
- ## Claude Code plugins
- Plugins for Claude Code only, installed from this repository's marketplace.
- <details>
- <summary><code>ste100</code>: answers in Simplified Technical English</summary>
- Makes the agent write to you in [ASD-STE100 Simplified Technical English](https://www.asd-ste100.org/): short
- sentences, active voice, one word per meaning, no idioms. It changes how the agent writes to you, not how it writes
- code.
- ```
- /plugin marketplace add CodelyTV/agent-harness
- /plugin install ste100@codely
- ```
- Then run `/config`, look for **Output Style**, and select `ASD-STE100`. More in its
- [README](.claude/plugins/ste100/README.md).
- </details>
- <details>
- <summary><code>ts7-lsp</code>: faster TypeScript code intelligence</summary>
- Go-to-definition, references and diagnostics through the
- [TypeScript 7 native compiler](https://devblogs.microsoft.com/typescript/announcing-typescript-7-0/) LSP. It indexes
- large monorepos roughly 10x faster and with far less CPU than the official `typescript-lsp` plugin.
- Requires a TypeScript 7+ `tsc` binary on `PATH` (e.g. `npm install -g typescript`).
- ```
- /plugin marketplace add CodelyTV/agent-harness
- /plugin install ts7-lsp@codely
- /plugin disable typescript-lsp@claude-plugins-official
- ```
- More in its [README](.claude/plugins/ts7-lsp/README.md).
- </details>
- ## Under the hood
- <details>
- <summary>One source for every agent</summary>
- Rules live in `AGENTS.md` and skills in `.agents/skills/`. Each agent expects them in a different path, so a `make`
- command creates the symlinks it needs. They are not committed, so run the ones you need after cloning:
- | Command | What it does |
- |--------------------------|-------------------------------------------------------------------------------------------------|
- | `make claude-symlinks` | Links `CLAUDE.md → AGENTS.md` in every directory with an `AGENTS.md`, and `.claude/skills` |
- | `make codex-symlinks` | Links `.codex/skills → .agents/skills` |
- | `make copilot-symlinks` | Links `.github/skills → .agents/skills` |
- | `make cursor-symlinks` | Links `.cursor/skills → .agents/skills` |
- | `make junie-symlinks` | Links `.junie/skills → .agents/skills` |
- | `make opencode-symlinks` | Links `.opencode/skills → .agents/skills` |
- </details>
- <details>
- <summary><code>export</code> blocked via hooks</summary>
- The `export` command can leak environment variables (tokens, secrets) if an agent runs it. Each agent has a
- pre-execution hook that blocks any shell command containing `export`:
- | Agent | Hook location | Mechanism |
- |-------------|------------------------------------------------------|--------------------------------------------------------------------------------------------------|
- | Claude Code | `.claude/settings.json` | Uses the `if` keyword with a glob pattern (`Bash(*export*)`) to match and block inline |
- | Cursor | `.cursor/hooks.json` + `hooks/block-export.sh` | Runs a shell script that parses the command via `jq` and exits with code 2 on match |
- | Copilot | `.github/hooks/hooks.json` + `hooks/block-export.sh` | Same approach as Cursor, adapted to Copilot's hook input format |
- | OpenCode | `.opencode/plugins/block-export.ts` | A native TypeScript plugin (`tool.execute.before`) that inspects the command and throws to block |
- </details>
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.