Oh My Pi CLI (subagent / task delegation)

What is it

This skill teaches your agent to use an agent from another vendor as a subagent.

Almost every coding agent has a headless mode. Headless mode runs the agent from the terminal with one command and returns the answer to stdout. This skill gives the command pattern for one such agent, and the protocol to transfer enough context with the task.

You stay the main agent. You keep the plan, the decisions and the conversation with the user. The subagent does one bounded task and reports back.

When to delegate

• The user asks for it. The user names another agent, or asks for a second opinion.
• Large or multi-step work. The task covers several files, phases or checkpoints.
• Code review with a fresh mind. Another agent reviews the work with a different lens and no memory of the decisions that produced it.
• Planning that needs different ideas. A second agent proposes options that you did not consider.
• An internal council. Several subagents answer the same question, and you compare the answers.
• Deep research. The other CLI has web search, web extraction or browser automation.
• Background investigation. A long task runs in the background while you continue in the foreground.

When not to delegate

• Small or single-step tasks. One or two edits, or a short explanation, are faster in this thread.
• Tight feedback loops. The user wants quick back-and-forth in one conversation.
• Secrets or policy-sensitive work. Do not pipe credentials into a subagent. Redact first.
• Context you already hold. Repeating the whole plan in a prompt adds no value.
• Low return on investment. If writing the handoff costs more than doing the edit, do the edit.
• Work that needs high precision and full judgement. A subagent without your context will break a task that depends on a detail only this session knows.

Delegation and context transfer protocol

An isolated subagent saves tokens, but it splits the story. The subagent does not see this conversation. It has no memory of the goal you agreed with the user, the decisions you already made, or the paths you agreed not to touch. A poor handoff causes misread tasks, wrong assumptions about the stack or the style, and wasted edits.

Treat every call as stateless delegation: the subagent keeps nothing between calls, so each command must carry its own state. A good handoff happens when you keep these practices in mind:

• Shared state handoff. The prompt carries the original goal, the decisions already made, and the explicit scope, to bridge the subagent's missing session history.
• Contextual preservation. The high-volume work stays in the isolated process, so the bulk never enters this conversation and no "split story" forms from missing history.
• Result reconciliation. What comes back is folded into this thread afterwards, so there is a single source of truth.

Write the prompt as a transfer of shared state, not as a title. Include all six fields:

Prefer sequential delegations with explicit carry-over. Parallel runs diverge unless every run gets the same briefing.

Run the subagent in the mode that does the work without asking for approval. Headless mode has nobody to answer a permission prompt. Any mode that stops to ask will stall, or return an empty answer with a success exit code. Each CLI names this mode differently — take the flag from the vendor card below.

Never call a subagent in plan mode. Many CLIs have a plan or read-only mode. That mode makes the other agent write a plan for itself, which is not what you asked for: you want its work or its answer, and you keep the planning. Plan mode also waits for someone to approve that plan, so the run comes back with nothing. If you want no file changes, keep the auto-approving mode and write "report only, no edits" in the Output field. Do not use plan mode to make a run read-only.

After the subagent returns:

• Report to the user. Give a short summary. Do not paste long logs unless the user asks.
• Reconcile the context. Record the decisions it made, the files it changed, and the open risks, so this session stays the single source of truth.

CLI failure modes

Every vendor has its own terminal rules. A headless run fails quietly more often than it fails loudly.

• Wrap the call in an external timeout. A headless CLI can stall before its own timeout starts.
• Exit code 0 is not success. If stdout is empty, treat the run as failed and read stderr. A permission the CLI could not ask for, and a prompt that never arrived, both look like success.
• Never carry a flag habit from one CLI to another. The same short flag means different things in different tools. In OpenCode, -p is --password: passing a prompt to it empties the message and the run waits on stdin forever. Confirm every flag against the CLI's own --help.
• An unrecognized-flag error means this skill is stale, not that the task is impossible. Read --help, continue with the flags that exist, and tell the user which line in this file is wrong.

Vendor card

Binary and prompt form

• Binary: omp
• Install: bun install -g @oh-my-pi/pi-coding-agent (the README recommends bun), or curl -fsSL https://omp.sh/install | sh, or brew install can1357/tap/omp (all from https://github.com/can1357/oh-my-pi).
• Prompt form: positional, gated by a boolean flag. -p/--print runs omp non-interactively (processes the prompt, streams the result to stdout, exits) but takes no value itself — the prompt text follows it positionally: omp -p "prompt". A bare omp opens the interactive TUI and never returns.

Headless and output flags

Approvals and permissions

Four tiers, set with --approval-mode <mode> (or --auto-approve/--yolo to force yolo):

For delegated/subagent runs specifically, the docs state: "Subagents run headless with tools.approvalMode: yolo so ordinary tier-based prompts do not stall them." A per-tool tools.approval.<tool> override still applies on top of the mode: deny blocks the tool, allow permits it, and prompt "cannot be satisfied headlessly and thus rejects the call" — a rejection, not a hang, is the documented failure signature. --plan-yolo forces read-only plan mode at start and auto-approves the plan — do not use it for delegation; it still plans instead of acting. No exit-code table is documented — treat empty stdout as failure.

Models and how to list them

Run omp models (default ls) for provider-grouped tables of every available model; omp models find <substring> filters, omp models refresh forces a re-fetch, and a bare provider name also filters (omp models openai-codex). --json gives machine-readable output.

--model <id-or-role> accepts an exact provider/modelId, a bare model ID, a fuzzy/substring match, or a glob scope (openai/*, *-mini*), each optionally suffixed :thinkingLevel (off|minimal|low|medium|high|xhigh|max). Roles (default, smol, slow, vision, plan, commit, tiny, task, advisor) resolve through settings.modelRoles; --smol, --slow, --plan <id> override specific roles directly. Use the web and the vendor's own docs for pricing and benchmarks.

Command pattern

omp -p "GOAL: [goal] | DECISIONS: [decisions] | SCOPE: [paths] | CONSTRAINTS: [constraints] | VERIFICATION: [test_command] | OUTPUT: [format]" \
  --yolo --mode json --model <id-or-role> --max-time 45m 2>&1

Prompt examples

• Implement: omp -p "GOAL: [goal] | SCOPE: [paths] | VERIFICATION: [test_command] | OUTPUT: files changed" --yolo --max-time 45m
• Review (report only): omp -p "GOAL: Review [scope] for [concerns] | CONSTRAINTS: do not edit any file | OUTPUT: findings with severity" --approval-mode always-ask
• Investigate: omp -p "GOAL: Map how [feature] works | SCOPE: [paths] | CONSTRAINTS: do not edit any file | OUTPUT: concise file:line map" --yolo

Docs and reference

• Flags, approval modes, models, skills, auth, paths: reference.md
• Vendor documentation: https://github.com/can1357/oh-my-pi/tree/main/docs (the vendor's own source; the omp.sh/docs site renders from it)
• Checked against the official documentation on 2026-11-05; not run against an installed binary. Latest release on npm: @oh-my-pi/pi-coding-agent 18.16.0. Confirm with omp --help before trusting a flag.