How do I install the Tao Boomerang skill?

Run npx skillmds@latest add cleanexpo/tao-boomerang in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Tao Boomerang safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Tao Boomerang?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Tao Boomerang free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Tao Boomerang?

cleanexpo (@cleanexpo) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Tao Boomerang by cleanexpo — Agent Skill

Name: Tao Boomerang
Author: cleanexpo

Tao Boomerang

Token-efficient one-shot dispatch with summary-only return. Caller asks a question, the worker replies with ONLY the answer — no preamble, no chain-of-thought, no markdown unless the answer requires it. Composes on `_run_claude_via_sdk`. Use when you need the answer, not the conversation.

cleanexpo Updated Sep 17, 2026

File contents

tao-boomerang

The "I just need the answer" pattern. Wraps _run_claude_via_sdk with a summary-only preamble + post-strip of conversational artefacts.

Public API

from app.server.tao_boomerang import boomerang, dispatch_one

# Single
result = await dispatch_one(
    question="What's the latest stable Python version?",
    timeout_s=120,
)
# result.summary, result.cost_usd, result.elapsed_s, result.error

# Parallel
batch = await boomerang(
    questions=["q1", "q2", "q3"],
    max_parallel=5,
    timeout_s=120,
)
# batch.results, batch.total_cost_usd, batch.all_succeeded

UNKNOWN: <reason> is a contract — the worker returns this when the question can't be answered without more context. Callers can branch on result.is_unknown.

When to trigger

Research subagent calls where the caller throws away the reasoning trace anyway.
Fact lookups that don't benefit from chain-of-thought (e.g. "current X price", "latest version of Y").
Parallel batch queries where the per-result summary is what feeds the next stage.

When NOT to use

Anything that needs reasoning visible to the caller — use _run_claude_via_sdk directly with default thinking mode.
Iterative goal pursuit — use tao-loop for that.
Tool-heavy workflows (file edits, web fetch, etc.) — boomerang disables extended thinking and discourages tool use to keep output tight.

Autoresearch envelope

Slot	Value
Single metric	tokens-returned vs tokens-if-conversational
Time budget	per-call timeout, default 120s
Constrained scope	`app/server/tao_boomerang.py`
Strategy/tactic split	caller frames question; module enforces summary-only
Kill-switch	inherits SDK timeout + asyncio.wait_for wrapper

CLI

python scripts/run_boomerang.py "Your question"
python scripts/run_boomerang.py "q1" "q2" "q3" --max-parallel 3
python scripts/run_boomerang.py --json "What is X?"

Composition

vs parallel-delegate (Agent tool): boomerang is leaner. Agent tool returns the full subagent transcript. Boomerang strips to one paragraph max. Use Agent for "explore + report"; use boomerang for "answer this one thing".
vs tao-loop: loop is iterative + judge-gated. Boomerang is one-shot. Don't compose them — if you need a goal-pursuit loop, go straight to tao-loop.
vs Margot deep_research: Margot does multi-source synthesis with citations. Boomerang is a single SDK call. Use Margot when the answer needs grounding in the corpus or web.

Behaviour invariants

Worker model: Sonnet (select_model("generator")). Never opus.
thinking="disabled" per dispatch — no extended thinking budget.
Summary is post-stripped: leading "Sure, " / "Here is the answer: ", trailing "Let me know if..." footers, single-block markdown fences. Conservative: only strips when the WHOLE response matches a pattern, never edits the answer body.
error populated on timeout / SDK error; result.summary empty in that case. Callers MUST check error before consuming summary.

---name: tao-boomerangdescription: Token-efficient one-shot dispatch with summary-only return. Caller asks a question, the worker replies with ONLY the answer — no preamble, no chain-of-thought, no markdown unless the answer requires it. Composes on `_run_claude_via_sdk`. Use when you need the answer, not the conversation.---# tao-boomerangThe "I just need the answer" pattern. Wraps `_run_claude_via_sdk` witha summary-only preamble + post-strip of conversational artefacts.## Public API```pythonfrom app.server.tao_boomerang import boomerang, dispatch_one# Singleresult = await dispatch_one( question="What's the latest stable Python version?", timeout_s=120,)# result.summary, result.cost_usd, result.elapsed_s, result.error# Parallelbatch = await boomerang( questions=["q1", "q2", "q3"], max_parallel=5, timeout_s=120,)# batch.results, batch.total_cost_usd, batch.all_succeeded````UNKNOWN: <reason>` is a contract — the worker returns this when thequestion can't be answered without more context. Callers can branchon `result.is_unknown`.## When to trigger- Research subagent calls where the caller throws away the reasoning trace anyway.- Fact lookups that don't benefit from chain-of-thought (e.g. "current X price", "latest version of Y").- Parallel batch queries where the per-result summary is what feeds the next stage.## When NOT to use- Anything that needs reasoning visible to the caller — use `_run_claude_via_sdk` directly with default thinking mode.- Iterative goal pursuit — use `tao-loop` for that.- Tool-heavy workflows (file edits, web fetch, etc.) — boomerang disables extended thinking and discourages tool use to keep output tight.## Autoresearch envelope| Slot | Value || --- | --- || Single metric | tokens-returned vs tokens-if-conversational || Time budget | per-call timeout, default 120s || Constrained scope | `app/server/tao_boomerang.py` || Strategy/tactic split | caller frames question; module enforces summary-only || Kill-switch | inherits SDK timeout + asyncio.wait_for wrapper |## CLI```bashpython scripts/run_boomerang.py "Your question"python scripts/run_boomerang.py "q1" "q2" "q3" --max-parallel 3python scripts/run_boomerang.py --json "What is X?"```## Composition- **vs parallel-delegate (Agent tool):** boomerang is leaner. Agent tool returns the full subagent transcript. Boomerang strips to one paragraph max. Use Agent for "explore + report"; use boomerang for "answer this one thing".- **vs tao-loop:** loop is iterative + judge-gated. Boomerang is one-shot. Don't compose them — if you need a goal-pursuit loop, go straight to tao-loop.- **vs Margot deep_research:** Margot does multi-source synthesis with citations. Boomerang is a single SDK call. Use Margot when the answer needs grounding in the corpus or web.## Behaviour invariants- Worker model: Sonnet (`select_model("generator")`). Never opus.- `thinking="disabled"` per dispatch — no extended thinking budget.- Summary is post-stripped: leading "Sure, " / "Here is the answer: ", trailing "Let me know if..." footers, single-block markdown fences. Conservative: only strips when the WHOLE response matches a pattern, never edits the answer body.- `error` populated on timeout / SDK error; `result.summary` empty in that case. Callers MUST check `error` before consuming summary.

Frequently asked questions