Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Retrospect External Babysitter Run

Locates babysitter processes and committed runs in an external public repo and writes an evidence-backed retrospective with concrete improvement suggestions.

AutomationAdvanced★ 1,797⑂ 106AI score 8/10Last updated: Sep 16, 2026

What it does

  • Picks a target repo from docs/repo-with-babysitter-processes.md and shallow-clones it with gh repo clone --depth 50.
  • Finds process files importing @a5c-ai/babysitter-sdk and .a5c/runs/<runId>/ artifacts (run.json, inputs.json, journal events, tasks, state/output.json); recovers historical runs from git history when .a5c/runs/ is gitignored.
  • Analyzes outcome, what went well, what went poorly (re-dispatched effects, long EFFECT_REQUESTED→RESOLVED gaps, looping breakpoints, crash error categories).
  • Reviews process quality: determinism and stable invocation keys, effect granularity, idempotency, breakpoint discipline, error surfacing, labels, reuse potential.
  • Writes a templated report to docs/retrospectives/<owner>-<name>/<runId>.md plus a log line so runs aren't retrospected twice, and routes generalizable insights to /babysitter:contrib.

Who it's for

  • Babysitter process authors who want to learn from how others use (and misuse) the framework.
  • Contributors filing upstream SDK/CLI/doc feedback with hard evidence.
  • Teams building an evidence-first postmortem habit around run journals.

Example uses

  1. "Retrospect on repo X's latest failed run" → traces the final journal events and thrown error into a structured report.
  2. "How are other people using babysitter processes, and what do they get wrong?" → picks a high-star catalog repo and performs a process-quality review.
  3. Repo with no committed runs → switches to a clearly labeled process-only retrospective focused on determinism and breakpoint design.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/a5c-ai/babysitter/HEAD/.claude/skills/retrospect-external-babysitter-run/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the .claude/skills/retrospect-external-babysitter-run folder from the GitHub repo a5c-ai/babysitter into my ~/.claude/skills/retrospect-external-babysitter-run/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone --depth 1 https://github.com/a5c-ai/babysitter.git /tmp/babysitter && mkdir -p ~/.claude/skills && cp -r /tmp/babysitter/.claude/skills/retrospect-external-babysitter-run ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal and clone the repo: git clone --depth 1 https://github.com/a5c-ai/babysitter.git /tmp/babysitter
  2. Create the skills folder: mkdir -p ~/.claude/skills
  3. Copy the skill: cp -r /tmp/babysitter/.claude/skills/retrospect-external-babysitter-run ~/.claude/skills/
  4. Install and authenticate GitHub CLI: gh auth login (required for cloning external repos).
  5. If docs/repo-with-babysitter-processes.md doesn't exist yet, run the catalog-babysitter-users skill first.
  6. Restart Claude Code, then ask: "review the external babysitter run in <owner>/<repo>".
View source on GitHub ↗License: MIT