CtrlK
BlogDocsLog inGet started
Tessl Logo

x-verify

Use when a long-running skill needs to check "am I done?" — runs the canonical completion cascade with mandatory fallback to prevent silent success claims

71

Quality

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

x-verify — Completion Cascade

Purpose

Single entry point for answering "am I done?" reliably. Every long-running x-skill dispatches here instead of running its own ad-hoc checks.

See ../x-shared/completion-cascade.md for the full cascade specification.

Role: verifier

This skill is a verifier. It reports completion status; it does not apply fixes.

x-verify MUST NOT:

  • Call Edit or Write — if fixes are needed, return findings and let the caller route to an executor
  • Call mutating Bash commands — only read-only verification (tests, lint, typecheck, git log)
  • Claim "done" when the verification cascade didn't actually complete

Execution

Run the cascade from ../x-shared/completion-cascade.md in order. Do not re-document the cascade here — that file is the single source of truth. If you are editing this file to change cascade logic, stop: edit completion-cascade.md instead.

High-level shape (pointer only, full detail in the canonical file):

  1. SCOPE GATE — un-tooled or docs-only invocation short-circuits to done
  2. ABORTEXPLICIT FAILUREVERIFICATIONMANDATORY FALLBACKHUMAN-APPROVAL

Verifier dispatch (step 4): call Agent tool with subagent_type: "oh-my-claudecode:code-reviewer". Claude-only fallback when OMC is unavailable: Agent tool with a generic review prompt (no subagent_type). See the canonical file for the fallback contract.

Output format

Return one of these verdicts:

verdict: done
reason: all-checks-passed
details:
  test: passed
  lint: clean
  typecheck: clean
  fallback: (not invoked)
verdict: failed
reason: test-failed
details:
  test: FAIL (3 failures)
  lint: clean
  typecheck: clean
  findings: [ ... ]
verdict: aborted
reason: user-abort           # or stagnation-option-D
details: { ... }
verdict: waiting-for-user
reason: stagnation-menu-open # or human-approval-needed
details: { ... }
menu: [A] alternative-A, [B] alternative-B, [C] alternative-C, [D] abort
verdict: needs-user-review
reason: all-verification-inconclusive
details:
  test: no-config
  lint: no-config
  typecheck: no-config
  fallback: uncertain (see findings)
  findings: [ ... ]
menu: [A] mark done, [B] re-verify, [C] abort

Rationale

Closes the "verification-before-completion skipped" compliance gap documented in feedback_xreview_compliance.md. The mandatory fallback prevents skills from claiming done when they have no actual verification signal.

Repository
quangtran88/x-skills
Last updated
Created

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.