Skip to content
SkillModel-led815 tokens in use

verification-before-completion

Requires fresh command output as evidence before your agent claims work is done, fixed or passing.

322,762 stars on the repository MIT Licenseobra/superpowers · skills/verification-before-completion
/plugin install superpowers@claude-plugins-official
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing verification-before-completion costs

43tokens idle
815tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter12/15
  • Structure17/25
  • Examples12/15
  • Output15/15
  • Size15/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity56
  • Maintenance68
  • Community64
  • Maturity36
  • Momentum40

Verification Before Completion sets one rule: no completion claims without fresh verification evidence. Before saying anything is done, the agent identifies the command that proves the claim, runs it in full, reads the output and exit code, and only then states the result with the evidence.

What You Get

  • A five-step gate: identify, run, read, verify, then claim
  • A table of claims, the evidence each needs and what falls short
  • A red-flag list of wording that implies success without a run
  • Patterns for tests, regression tests, builds, requirements and delegated work

Common Use Cases

  • Before committing or opening a pull request
  • After a subagent reports success
  • Before marking a plan task complete

What Makes It Unique

It covers wording as well as behavior: paraphrases, synonyms and expressions of satisfaction all count as claims that need evidence. It also includes a red-green check for regression tests, requiring the test to fail with the fix reverted.

Under The Hood

Instructions only. The skill ships no scripts and calls nothing itself. It sends the agent to run whatever verification command fits the claim, such as a test runner, linter or build, and to check the version control diff after delegated work.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers