skills/context/SKILL.md
Switch the active credential profile for ronan-skills. Three-tier resolver picks the right `~/.claude/.env.<context>` file based on `$RO_CONTEXT` env var, manual override, `.ro-context.local` / `.ro-context` files in the cwd or any ancestor, cwd-glob rules in `~/.claude/contexts.json`, or the default. Skills source via `$(ro context env)` to get a cwd-aware credential file. Use when the user wants to switch credentials between contexts (personal / work / per-client), set up a new context, or debug which context resolved.
npx skillsauth add RonanCodes/ronan-skills contextInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
ro context is the meta-CLI all other ronan-skills lean on for credential discovery. The actual script lives at skills/context/scripts/ro (in this repo) and is symlinked into ~/.local/bin/ro so it's invokable from any terminal as ro context ….
This SKILL.md is the user-facing documentation. The script's --help is the canonical command reference.
| Question | Command |
|---|---|
| What context am I in right now? | ro context show |
| Why this context (which rule fired)? | ro context where |
| Which env file did the resolver pick? | ro context env |
| Which contexts exist? | ro context list |
| Switch context for the rest of my session | ro context use <name> |
| Stop pinning, go back to cwd-resolution | ro context auto |
| First-time setup | ro context init |
| Add a cwd-glob auto-rule | ro context add-rule '~/Github-Acme/**' acme |
$RO_CONTEXT env var — one-shot per shell, never written to disk.contexts.json:active — manual override set by ro context use. Sticky until cleared by ro context auto..ro-context.local in cwd or any ancestor — gitignored contributor override..ro-context in cwd or any ancestor — committed, repo's declared context. Recommended default for any repo with a stable owner/account.contexts.json — e.g. ~/Github-Acme/** → acme.default in contexts.json — fallback.The "ancestor" walk goes up to $HOME and stops there, never above.
Every skill that needs API tokens does:
set -a; source "$(ro context env)"; set +a
$(ro context env) returns the absolute path of the env file the resolver picked. For single-context users that's always ~/.claude/.env.personal; for multi-context users it varies per cwd.
Backwards compat: ~/.claude/.env is still a symlink to whichever context was last ro context use'd. Skills that haven't migrated to $(ro context env) keep working unchanged.
Everything under ~/.claude/ (no split with ~/.config/):
~/.claude/
contexts.json # { default, active, rules: [[glob, name], ...] }
.env # symlink, kept for backwards compat
.env.personal # real file per context (mode 600)
.env.<other> # one per profile
# 1. Copy the personal template as a starting point
cp ~/.claude/.env.personal ~/.claude/.env.acme
chmod 600 ~/.claude/.env.acme
# …edit with the client's credentials…
# 2. Pin a repo to its context (preferred — committed, no global config needed)
echo acme > /path/to/acme/repo/.ro-context
# OR add a cwd-glob auto-rule
ro context add-rule '~/Github-Acme/**' acme
| If you… | Resolution |
|---|---|
| Forget to switch context, open a terminal in a Simplicity repo | .ro-context resolves automatically |
| Clone a new repo from a client without thinking | cwd-glob (~/Github-<Client>/**) catches it |
| Have overlapping repos in ~/Dev/ (work + personal side-by-side) | Each repo's own .ro-context wins over the glob |
| Want to test cross-account scripts | RO_CONTEXT=other-account env var for one shell |
| Want to "stay in" one context for a long-running session | ro context use <name> until ro context auto |
| Have a long-running process (dev server, Claude session) | The process holds whatever it was started with — symlink swap doesn't poke it. Restart needed |
[[ro-context-env-profiles]] (LLM wiki research vault) — the design rationale + pull-out checklist~/.claude/contexts.json — the user's actual rules + defaultscripts/ro (this skill) — the implementation, ~320 lines of bash, zero dependencies beyond bash + jq + awk + sedtesting
--- name: linear-pipeline description: The Fable orchestrator for a single dispatched Linear ticket. Holds almost no context itself; it receives `--issue <ID> --detached`, decides the stage sequence, and fans out a sub-agent per stage, passing forward only each stage's artifact (never re-derived, never inlined into its own context). Step zero, before any planning or stage routing, is a boundary triage against `canon/security-boundary.md` (#199): a match tags Ronan Connolly and stops the run, no
development
--- name: in-your-face description: Capture a chat-only answer into a durable artifact (markdown + HTML, PDF when cheap) and launch it automatically so the user cannot miss it. Use when user says "in your face", "don't let me lose this", "save that answer", "make that durable", or right after answering a substantive side question (a recipe, comparison, how-to, or generated prompt) that would otherwise die with the context. category: workflow argument-hint: [--no-open] [--vault <short>] [hint of
tools
One-shot headless OpenAI Codex CLI calls for background/admin AI tasks — summaries, classification, extraction, admin glue. The default engine for anything that runs AI constantly in the background (daemon-driven, per-event), because it bills the flat ChatGPT subscription instead of Claude usage or per-token API spend, and it keeps working while Claude is rate-limited. NEVER for coding — coding stays Claude. Use when a skill or daemon needs a cheap always-on AI call, when the user says "use codex", "ask codex", "codex as backup", or when building a background summarizer/classifier into a listener or loop. Reads auth from ~/.codex/auth.json (ChatGPT account, no API key).
research
Turn a warranty rejection, repair quote, or RMA email into a cited decision brief — legal read (NL/EU consumer law), is the part user-serviceable, live part and new-unit prices, repair-vs-DIY-vs-new economics, before-you-send-it checklist, deadlines. Use when the user pastes or screenshots a repair quote, warranty rejection, "not covered" email, onderzoekskosten fee, or asks "should I repair or replace this".