feat: add mulesoft-dx skills export (39 skills)

[spezzirriemiliano]

Mirrors 39 Claude Code skills from mulesoft-dx into sf-skills' flat Agent Skills format. Prose skills are copied near-verbatim; jtbd skills have their api:/operationId YAML step blocks flattened into prose in SKILL.md, with full parameter sourcing and output mapping preserved in a companion references/api-steps.md.

Generated by the mulesoft-dx-sf-skills-export pipeline.

References: Contributing guide · Skill authoring guide · Agent Skills spec

What changed

Adds 39 skills to the root catalog, mirrored from mulesoft-dx's Claude Code skills:

• 27 prose skills — copied near-verbatim, renamed to avoid catalog collisions: mule-development/* skills get a mule-development- prefix (they don't self-prefix in the source repo), omni-gateway/* skills keep their existing self-prefix as-is, and platform-assistant is unprefixed (single skill, no collision risk).
• 12 jtbd (job-to-be-done) skills — these have machine-validated api:/operationId YAML step blocks in the source repo (checked against live Anypoint OpenAPI specs). Each was flattened into plain prose in SKILL.md for this catalog's format, with the full parameter-sourcing and output-mapping chain preserved in a companion references/api-steps.md (mirrors the SKILL.md + references/*.md pattern already used here, e.g. dx-apexguru-scan).

Every jtbd SKILL.md includes a back-pointer note to its source in mulesoft-dx, since mulesoft-dx remains the source of truth (its version is machine-validated; this mirror is not).

Why

mulesoft-dx's skills aren't discoverable from this catalog today. Rather than migrating them out of mulesoft-dx (which would break validate_jtbd.py's guarantee that API references resolve against live specs), we're dual-publishing: mulesoft-dx keeps authoring/validating as the source of truth, and this is the export of everything currently in it into this catalog's flat format.

This mirror is generated by a small pipeline (mulesoft-dx-sf-skills-export) that treats mulesoft-dx as read-only, tracks a content-hash manifest per skill, and re-exports only what's new/changed on demand.

Notes

• This is a batch mirror, not 39 individually-authored contributions — each skill was already reviewed/maintained in mulesoft-dx; this PR just brings the existing set into this catalog for the first time. The manual "Skills" checklist below is written for single-skill submissions, so I'm not checking it box-by-box per skill — happy to address specific feedback on any of the 39 individually if reviewers spot something.
• Going forward, changes will come as smaller follow-up PRs from the same pipeline whenever mulesoft-dx's skills change (a release-flow hook there now prompts to check for drift and re-export).
• Open to feedback on the flatten approach for jtbd skills (prose summary + references/api-steps.md) if there's a preferred convention here for machine-derived multi-file skills.

---

Skills

Manual checklist

Description quality

• Describes what the skill does and the expected output
• Includes relevant Salesforce domain keywords (Apex, LWC, SOQL, metadata types, etc.)
• Trigger phrases are specific enough for Vibes to select this skill reliably

Instructions

• Clear goal statement
• Step-by-step workflow
• Validation rules for generated output
• Defined output / artifact

Context efficiency

• Core instructions are concise — supporting material lives in templates/, examples/, or docs/ subdirectories
• No unnecessary background explanation in the body

[salesforce-cla]

Thanks for the contribution! Unfortunately we can't verify the commit author(s): Emiliano Spezzirri <e***@s***.com>. One possible solution is to add that email to your GitHub account. Alternatively you can change your commits to another email and force push the change. After getting your commits associated with your GitHub account, refresh the status of this Pull Request.

[salesforce-cla]

Thanks for the contribution! Before we can merge this, we need @spezzirriemiliano to sign the Salesforce Inc. Contributor License Agreement.