Skip to content

skill: document the three ways to run ilo from an agent#155

Merged
danieljohnmorris merged 1 commit into
mainfrom
feature/skill-runtime-modes
May 11, 2026
Merged

skill: document the three ways to run ilo from an agent#155
danieljohnmorris merged 1 commit into
mainfrom
feature/skill-runtime-modes

Conversation

@danieljohnmorris
Copy link
Copy Markdown
Collaborator

Summary

  • Replace the pi-specific line in `skills/ilo/SKILL.md` with a generic three-tier section: named tools (e.g. `ilo_run` / `ilo_repl`), `ilo serv` JSON loop, one-shot shell

Why

One skill, one source of truth, used by every agent. The earlier line named pi explicitly; this version is agent-neutral and forward-compatible. `ilo serv` is a binary feature so it belongs in the general skill regardless of which agent is reading.

Test plan

  • After merge, cut a release; the Claude marketplace plugin and `pi-ilo-lang` both ship the updated skill
  • Sync local `~/.claude/skills/ilo/SKILL.md` to match

replace the pi-specific note with a generic three-tier section that
applies to any agent reading the skill:

1. named tools (ilo_run / ilo_repl) - when the agent exposes them
2. ilo serv - long-lived JSON request/response loop, agent-agnostic
3. one-shot shell - the universal fallback

named tools today means pi-ilo-lang inside pi, but the phrasing is
forward-compatible with any future agent that registers the same
tools (e.g. via an MCP server). ilo serv is a binary feature so it
belongs in the general skill regardless of which agent is reading.

agents that lack the tools simply read past option 1 and use option
2 or 3, so the addition is harmless for claude code, codex, cursor,
or anyone else.
@danieljohnmorris danieljohnmorris merged commit 3eebfd6 into main May 11, 2026
4 checks passed
@danieljohnmorris danieljohnmorris deleted the feature/skill-runtime-modes branch May 11, 2026 15:47
@codecov
Copy link
Copy Markdown

codecov Bot commented May 11, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ All tests successful. No failed tests found.

📢 Thoughts on this report? Let us know!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant