Babysitter verification-before-completion

Evidence requirement enforcement ensuring all claims are backed by logs, test results, or exit codes. Zero = success, non-zero = failure. No guessing allowed.

install

source · Clone the upstream repo

git clone https://github.com/a5c-ai/babysitter

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/a5c-ai/babysitter "$T" && mkdir -p ~/.claude/skills && cp -r "$T/library/methodologies/cc10x/skills/verification-before-completion" ~/.claude/skills/a5c-ai-babysitter-verification-before-completion && rm -rf "$T"

manifest: library/methodologies/cc10x/skills/verification-before-completion/SKILL.md

Verification Before Completion

Overview

Enforces the evidence-first principle: every claim of success, fix, or completion must be backed by concrete evidence. Exit codes, test output, and logs are the only acceptable proof.

Evidence Types

Exit codes: zero = success, non-zero = failure
Test output: pass/fail counts, coverage percentages
Logs: error messages, stack traces, resolution confirmation
Build output: compilation success/failure with timestamps

Rules

Never claim "should work" without evidence
Never claim success without exit code 0
Never claim a bug is fixed without reproduction failure
Always capture and report exit codes
Always run the full test suite, not just targeted tests
Use timeout guards (
```
timeout 60s
```
) to prevent hanging

When to Use

Before completing any BUILD workflow step
Before claiming a DEBUG fix is verified
Before approving a REVIEW result
Before marking any task as done

Agents Used

```
integration-verifier
```
(primary consumer)
```
component-builder
```
(TDD evidence)
```
bug-investigator
```
(fix evidence)