Share AI config CLI

Installs command, rule, and skill sources from ~/.config/dryai by default into Copilot and Cursor targets.

Pass --config-root <path> to read configs from a different root such as ./config.

Pass --output-root <path> to write generated output somewhere other than your home directory.

--test is a shortcut for --output-root ./output-test, and if both are provided, --output-root wins.

Input

Input config files live under the selected input root:

• commands
• rules
• skills

Live output is written to:

• ~/.copilot/prompts
• ~/.copilot/instructions
• ~/.copilot/skills
• ~/.cursor/rules
• ~/.cursor/skills

Example Configs

One input root can contain all three source types:

~/.config/dryai/
├── commands/
│   └── gen-commit-msg.md
├── rules/
│   └── say-yes-captain.md
└── skills/
    └── review-helper/
        └── SKILL.md

Example Rule

Rules are markdown files under rules/. dryai recognizes these rule frontmatter fields:

• description
• copilot.applyTo
• cursor.alwaysApply
• cursor.globs

cursor.globs should be provided as one comma-separated glob string.

---
description: Reply with "Yes, Captain!" before answering when the user says "Make it so" or "Engage".
copilot:
  applyTo: '**/*.tsx, **/*.ts, src/**/*.ts, src/**/*.tsx, src/**/*.js, src/**/*.jsx'
cursor:
  alwaysApply: false
  globs: '**/*.tsx, **/*.ts, src/**/*.ts, src/**/*.tsx, src/**/*.js, src/**/*.jsx'
---

# Say Yes Captain

When the user says "Make it so" or "Engage", start your response with "Yes, Captain!".

Example Command

Commands are markdown files under commands/. dryai recognizes these command frontmatter fields:

• name
• description
• cursor.disable-model-invocation

---
name: gen-commit-msg
description: Generate a conventional commit message from the current staged git diff.
cursor:
  disable-model-invocation: true
---

# Generate Commit Message

Read the staged diff and produce a conventional commit message with a concise subject and optional body.

Example Skill

Skills live in directories under skills/. The directory is copied as-is into the Copilot and Cursor skills targets.

Unlike commands and rules, dryai does not define or validate a fixed skill frontmatter schema. Skill files are passed through unchanged, so the allowed frontmatter fields depend on the skill format expected by the target editor or agent.

skills/
└── review-helper/
    └── SKILL.md

---
name: review-helper
description: Use this skill when the user asks for a code review, PR review, or wants bugs and risks called out first.
---

# Review Helper

Focus on findings first.

- Identify bugs, regressions, and missing tests before summarizing.
- Keep the overview brief unless the user asks for a deeper walkthrough.

Commands

dryai install
dryai skills list                                 List local skills
dryai skills add [options] <repo>                 Add managed skills from a remote repository
dryai skills remove <name>                        Remove a managed skill
dryai skills rehash <name>                        Refresh stored file hashes for one managed skill
dryai skills rehash-all                           Refresh stored file hashes for all managed skills
dryai skills update [options] <name>              Update a managed skill from its tracked source
dryai skills update-all [options]                 Update all managed skills from their tracked sources

<repo> may be a full git remote URL or a GitHub owner/repo shorthand such as anthropics/skills.

Managed Skills

Imported skills are copied into config/skills/<name>/ and tracked in skills.lock.json.

When you run skills commands, local skill directories and skills.lock.json are read from and written to the selected config root.

skills add requires at least one --skill <name> value. Each requested skill is always resolved from <repo root>/skills/<name>.

Use --as <name> to choose a different local managed skill name when importing exactly one skill.

Examples:

dryai skills add anthropics/skills --skill skill-creator
dryai skills add vercel-labs/agent-skills --skill pr-review commit
dryai skills add https://github.com/vercel-labs/agent-skills.git --skill pr-review commit

By default, imports track the requested ref. With no --ref, that means the remote default branch HEAD is tracked. Use --pin to store the currently resolved commit instead, so later skills update operations stay pinned to that commit.

The lockfile records:

• the local skill name
• the source repository
• the source path within that repository
• the requested git ref, when one was provided
• the resolved commit that was imported
• the last installed content hash for each file in the managed skill directory

skills update and skills update-all re-fetch the tracked repository snapshot and replace the local copied skill directory.

Before replacing a managed skill, dryai compares the current local files against the hashes stored in skills.lock.json. If any file was added, removed, or edited locally, the update is skipped and a warning is printed so you do not lose your customizations by accident.

For skills imported before hash tracking existed, use rehash to store hashes from the current local directory without fetching from the remote source:

dryai skills rehash review-helper
dryai skills rehash-all

Use --force to intentionally overwrite local edits:

dryai skills update review-helper --force
dryai skills update-all --force

skills remove deletes the local copied skill directory and removes its lockfile entry.

skills list reports local skill directories, annotating managed entries from the lockfile and flagging managed entries whose local directory is missing.

skills rehash updates the stored file hashes for one managed skill using its current local contents. skills rehash-all does the same for every managed skill and skips any managed entry whose local directory is missing.

Development

For development, use pnpm dev to rebuild into dest/ on change and pnpm dev:dryai --test install to run the built CLI.

Run pnpm run setup:editor after installing dependencies if you want the Effect language service workspace patch applied locally.

pnpm run setup:editor
pnpm run build
pnpm run dev
pnpm run test
pnpm run test:watch

pnpm dev:dryai install
pnpm dev:dryai --test install
pnpm dev:dryai --output-root ./tmp/install-root install
pnpm dev:dryai --config-root ./config install

CI and Release

• On pull request open or update

  • Run CI validation with build, test, and npm pack --dry-run.

• On changes landing on main

  • Run the same CI validation with build, test, and npm pack --dry-run.

• On v* tag pushed to main, the release workflow will:

  • Verify the tag matches the checked-in package.json version.
  • Verify the tagged commit is on main.
  • Build and test the CLI.
  • Create a tarball with npm pack.
  • Create or update the matching GitHub Release.
  • Upload the tarball as a release asset.

• After the release workflow succeeds, the publish workflow will:

  • Download the tarball artifact produced by the release workflow.
  • Publish that exact tarball to npm using npm trusted publishing.

Example release flow:

git tag v0.1.0
git push origin v0.1.0

Install from the release tarball with:

npm install -g https://github.com/willmruzek/share-ai-config/releases/download/v0.1.0/share-ai-config-0.1.0.tgz