Skip to main content
Use this as the base when creating a reusable DealDash skill.

​
Canonical Draft

See /docs/agents/DEALDASH_AGENT_SKILL.md.

​
Required Workflow In Skill

  1. read /AGENTS.md
  2. read change safety protocol
  3. run baseline checks
  4. implement scoped change
  5. update docs (/docs and /mintlify-docs)
  6. run verification checks
  • call out high-risk paths explicitly
  • require evidence-based handoff
  • enforce typed/validated boundaries for input data
  • require rollback note for risky changes

​
Why This Matters

A stable skill definition makes multi-agent execution predictable, reduces regressions, and keeps documentation quality from decaying over time.