agentleFS
Sign inStart free
gufrancogufranco/claude-engineering-rules/skills/aside/SKILL.md

aside

Skill3 starsChanged 31 days agoMITView on GitHub

Mid-task side question handler. Freezes the active task state, answers a side question without state mutation, then resumes the task exactly where it stopped. Lighter than /checkpoint save. Use when user says "aside", "quick question", "side question", "while we're at it", "before you continue", or interrupts mid-implementation with an unrelated question. For new scope, use /plan; for a retro, use /retro; for a true context switch, use /checkpoint save.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-only investigation is enough to answer

    line 21
  2. Distinctive:

    Do NOT use when:

    line 23
  3. Distinctive:

    Answer: Read-only investigation (Read, Grep, Glob, ToolSearch) followed by a direct answer.

    line 32
  4. Distinctive:

    Confirm scope unchanged: Ask whether the answer changes anything about the active task.

    line 33
  5. Distinctive:

    No write operations occurred between freeze and resume

    line 57

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Sidebar conversation that preserves the in-flight task. Read-only by design: no file edits, no shell commands. Answer, then return to the work that was interrupted.

Overview

Mid-task interruptions are a documented source of dropped work. The common failure modes are forgetting where work stopped, mixing the side answer's state into the main task, and silently switching scope without the user noticing. This skill defines an answer-only protocol: freeze, answer, resume.

When to Use

  • User asks a clarifying question while implementation is in progress
  • User asks "before you continue, what about X?" or "quick question"
  • A side topic arises that does not change the active task's scope
  • Read-only investigation is enough to answer

Do NOT use when:

  • The interruption changes the active task's scope (use /plan instead)
  • The interruption ends the task (use /checkpoint save instead)
  • The question requires editing files (use the normal flow, not aside)
  • The user is asking for a retrospective (use /retro)

Process

  1. Freeze. State the active task in one sentence and the immediate next step that was about to execute. Do not edit, do not run shell commands.
  2. Answer. Read-only investigation (Read, Grep, Glob, ToolSearch) followed by a direct answer. Keep the answer scoped to the question.
  3. Confirm scope unchanged. Ask whether the answer changes anything about the active task. If yes, exit aside and surface the scope change explicitly.
  4. Resume. Restate the active task and the immediate next step, then continue exactly where step 1 froze.

Common Rationalizations

  • "I will just make a small edit to demonstrate": no edits in aside. If demonstration requires edits, the side question is not a side question.
  • "I will skip the freeze step, it is obvious": the freeze IS the value. Skipping it is how mid-task state gets lost.
  • "I can answer this without reading anything": aside still requires verification per CLAUDE.md Anti-Hallucination. Read the file before answering.
  • "The user will remember where we were": the user does not need to. The skill does that work.

Red Flags

  • About to call Write, Edit, MultiEdit, or Bash while in aside
  • About to answer without naming the active task first
  • About to resume without restating the next step
  • Side answer that turns into a sub-implementation
  • More than one round-trip of aside without re-confirming the active task is unchanged

Verification

  • Active task is named in writing before the answer
  • Answer cites a file path or grep result (no fabrication)
  • Scope-change confirmation question was asked at step 3
  • Resume statement repeats the immediate next step verbatim
  • No write operations occurred between freeze and resume

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free