Skip to content

Add five agent skills for governed changes: undo, inbox, time off, rewards, WQL analyst - #24

Merged
Ekwuno merged 1 commit into
Workday:mainfrom
Daksh-Aneja-Projects:add-governed-agent-skills
Nov 11, 2026
Merged

Ekwuno merged 1 commit into
Workday:mainfrom
Daksh-Aneja-Projects:add-governed-agent-skills

Conversation

@Daksh-Aneja-Projects

Copy link
Copy Markdown
Contributor

What this PR adds or changes

Five agent skills for agents that act on Workday data with a person in the loop. Each one names the exact
Workday operations it relies on, treats every change as a proposal that a person approves, and lists what the
agent must refuse.

  • business-process-undo-agent-skill: read an event, then prepare a rescind (completed) or cancel (in progress)
  • inbox-decisions-agent-skill: read the inbox, explain a step, prepare approve, deny or send back
  • time-off-request-agent-skill: read balances and eligible types, prepare day-by-day time off
  • rewards-proposal-agent-skill: prepare a one-time payment or base pay change, amount always set by a person
  • wql-workforce-analyst-agent-skill: answer a workforce question with one read-only WQL SELECT

Checklist for new or changed examples

  • The example lives entirely in its own folder under examples/
  • node scripts/validate-examples.mjs --check passes (valid example.json, README present, index table in sync)
  • The README says what the artifact is, how to use it, and what to change before deploying it (the "Before you deploy" section)
  • No credentials, tenant names, or real personal data anywhere in the folder
  • node scripts/audit-examples.mjs --changed passes, or the findings it reports are explained below

Anything reviewers should know?

  • The audit was run with --skip-arcane (hub rules only); the folders hold markdown and JSON, no Extend source.
  • The skills come from an agent layer I built on a developer tenant. Reads and the sign-in ran live there. Not
    every write has been run to completion in a tenant, and each README says so under "Before you deploy".
  • Happy to split this into one PR per skill if that is easier to review.

🤖 Generated with Claude Code

Undo, inbox decisions, time off, rewards and a WQL workforce analyst.
Each skill names the Workday operations it relies on, treats every
change as a proposal a person approves, and lists what to refuse.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
@github-actions

Copy link
Copy Markdown

Example audit

Audited examples/business-process-undo-agent-skill, examples/inbox-decisions-agent-skill, examples/rewards-proposal-agent-skill, examples/time-off-request-agent-skill, examples/wql-workforce-analyst-agent-skill. No findings. Thank you for a clean submission.

Run it yourself: ./scripts/install-arcane.sh once, then node scripts/audit-examples.mjs --changed. Rule explanations and fixes: docs/EXAMPLE_BEST_PRACTICES.md.

@Ekwuno Ekwuno left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for contributing @Daksh-Aneja-Projects

@Ekwuno
Ekwuno merged commit de997cc into Workday:main Nov 11, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants