Subagent fanout
Skill pipipip169/fable5-handoff/sources/adamentwistle-fable-skills/subagent-fanout
Delegate broad searches and independent subtasks to subagents to keep the main context clean. Use when a question requires sweeping many files or directories, when several independent investigations could run in parallel, or when raw tool output would flood the context you need for the real work.From its SKILL.md
npx -y skills add pipipip169/fable5-handoff --skill subagent-fanoutAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.7 KB, 314 tokens by cl100k_base, as published. Nobody here has run it
subagent-fanout
Context spent on file dumps is context unavailable for reasoning. Spend subagent context on search; keep main context for synthesis and edits.
- Fan out when the answer spans many files: "where is auth handled", "find every caller of X", "which configs mention Y". Launch an Explore/general-purpose agent with a precise question and keep only its conclusion. Reading twelve files inline to answer one question is the anti-pattern.
- Parallelize independent work: three unrelated investigations = three agents in one message, not a sequence. Wall-clock and context both win.
- Write real briefs. A subagent has none of your context: state the goal, the constraints, what a good answer looks like, and what to return ("list of file:line + one-line role each" beats "look into auth"). Vague briefs return vague essays.
- Don't delegate the single lookup. If you know the file or symbol, one Grep/Read is faster than an agent round-trip. Delegation pays when breadth is the problem.
- Don't duplicate delegated work — once an agent is searching, don't run the same search inline while waiting.
- Trust but verify at the point of use: an agent's claim about a signature or behavior gets re-checked against the actual file before you edit based on it (see ground-before-edit).
Heuristic: if answering would mean reading more than ~4 files just to locate something, fan out.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.