Skip to content

README.md

Latest commit

 

History

History
83 lines (56 loc) · 3.11 KB

README.md

File metadata and controls

83 lines (56 loc) · 3.11 KB
Raw

revcat skills

Agent Skills for revcat. Each skill is a folder containing a SKILL.md with YAML frontmatter (name + description are required). Drop them into your agent's skills directory and the agent picks the right one based on the description.

What's here

  • revcat-getting-started/ — first-time install + auth (browser OAuth) + revcat init for per-repo project context + command map. Triggers when a user wants to start using revcat.
  • revcat-commands/ — real syntax + examples for every subcommand including init. Triggers when constructing a command.
  • revcat-troubleshooting/ — common errors and fixes (401, no profile, no project_id, v0.3-to-v0.4 upgrade paths, toml/local mismatch, Test Store quirks, v1-only endpoints, dashboard-only operations). Triggers on a failed command or error message.
  • revcat-storefront-debug/ — 7-step diagnostic flow for "the SDK sees 0 packages from my offering." Covers the Test Store price gotcha, product/store binding checks, and the v1 /subscribers/{id}/offerings verify step. Triggers on "0 packages", "fetchOfferings empty", and similar SDK-side surprises.

Install via skills.sh (recommended)

skills.sh is a registry for Agent Skills. The CLI auto-detects your agent (Claude Code, Cursor, Codex, etc.) and writes to the right config dir.

All four revcat skills:

npx skills add akshitkrnagpal/revcat

A single skill from the bundle:

npx skills add akshitkrnagpal/revcat/revcat-getting-started
npx skills add akshitkrnagpal/revcat/revcat-commands
npx skills add akshitkrnagpal/revcat/revcat-troubleshooting
npx skills add akshitkrnagpal/revcat/revcat-storefront-debug

Browse on the web: https://skills.sh/akshitkrnagpal/revcat.

Restart your agent (or open a new session) and the skills are live.

Install manually

If you can't or don't want to use skills.sh, copy the folders directly.

Claude Code

User-level (every project):

mkdir -p ~/.claude/skills
cp -R skills/revcat-* ~/.claude/skills/

Project-level (this repo only):

mkdir -p .claude/skills
cp -R skills/revcat-* .claude/skills/

Cursor

Cursor reads project-scoped rules from .cursor/rules/:

mkdir -p .cursor/rules
for s in skills/revcat-*; do
  name=$(basename "$s")
  cp "$s/SKILL.md" ".cursor/rules/$name.md"
done

Codex

Codex CLI reads from ~/.codex/skills/:

mkdir -p ~/.codex/skills
cp -R skills/revcat-* ~/.codex/skills/

If your agent uses a different path, follow its local install docs — the SKILL.md format is portable.

Authoring guidelines

  • Description should be trigger-focused ("Use when...") and front-load key terms. Keep under ~200 chars where possible.
  • Body should be concise and example-heavy. Reference real commands — if you're unsure, check revcat <group> --help or the docs.
  • Don't invent flags. If something is genuinely TBD, leave a TODO: note for the maintainer.
  • After breaking changes (e.g., the v0.4 OAuth-only rewrite), do a sweep across all skills — the troubleshooting skill in particular accumulates legacy error strings users will still search for.