Executes these capabilities
Inside this skill
The internal capability tree shows how this skill gets its work done.
SValidation Checklist
- criteria decomposition
- evidence collection
- pass/fail reporting
The contract
Input
- Acceptance criteria from the task brief
- The deliverable to check
Output
- Checklist with pass/fail and evidence per item
- Fix list for failures
Deliverables
- Completed checklist artifact
Use cases
- Final review before handoff
- Gate between agent and human approval
- Regression checks on updates
Dependencies and replacements
Dependencies
- A brief with explicit acceptance criteria
Required tools
- Browser or CLI tools as needed
Required permissions
- Access to the deliverable being checked
Compatible with
- visual-review
- link-checker
Alternatives
- Manual review without a checklist
- CI pipeline checks
Replaceable components
- The checklist format
- The evidence tools
Reliability
Human checkpoints
- Review failed items before accepting the work
Limitations
- Only as good as the acceptance criteria it checks
- Cannot prove quality, only criteria compliance
Failure modes
- Criteria written as feelings instead of checks
- Marking pass without collecting evidence
- Skipping items that are hard to test
Recovery steps
- Rewrite vague criteria as observable checks
- Add evidence screenshots or logs for each pass
- Test hard items first
Validation methods
- Another agent re-runs the checklist and gets the same results
Sources
Entry checked on Jul 31, 2026. Verification status reflects what AgentSkillTree has independently run, not the skill's own claims.
When to use
Use at the end of any task where “done” needs a definition. The checklist is the artifact a human reviews before accepting agent work.
How it works
Take each acceptance criterion and rewrite it as an observable yes/no check. Run the checks, collect evidence for each pass, and list fixes for each failure. Re-run until the checklist is clean.