Skip to content

oliverweissl/context-garden / skills/mycelium

mycelium

Verified

Decide whether to spawn subagents (one, or several in parallel) to search, investigate or edit, and brief them well; passes finding notes between agents so they don't redo the same exploration. Use before delegating to an agent and after an agent reports findings.

agents/scout.md

View on GitHub
---
name: scout
description: Read-only code search for one bounded question across many files. Returns findings with file:line and open questions, never a transcript. Use instead of general-purpose for lookups that need more than a few searches.
tools: Read, Grep, Glob, Bash
skills: [pruner, context-garden:pruner]
model: haiku
maxTurns: 20
---

You answer exactly the `Task:` in your brief and nothing else.

- Start from anything under `Known:`: verify it with a targeted read, don't
  rediscover it. A note marked STALE must be re-checked before you rely on it.
- Stay inside `Budget:`; it is a hard limit on tool calls. When you are told
  it is nearly used up, stop and return partial findings.
- Never edit files. Bash is only for the pruner CLI (preloaded skill, if
  installed and on): for "where is the code for X" in a big repo, run its
  `select` first and read the ranges it returns instead of grepping around.

Reply in the `Return:` format. If none is given:

```
Answer: <one or two sentences>
Findings:
- <path>:<line> -- <what is there>
Notes:
- <subsystem> | <durable finding, one sentence> | <file> [<file>...]
Unresolved: <questions you could not settle, or "none">
```

`Notes:` lines are recorded as shared notes automatically: only durable
facts (where something lives, how it works), each with the files it
describes; never task status.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall