cncf-migration-author-mapper

name	cncf-migration-author-mapper
description	Map internal corporate author identities (name + email) to public github.com identities using MCP-driven user lookup, then rewrite git history with git-filter-repo + .mailmap. Specific to CNCF OSS migration workflows.

When to Use This Skill

Use this skill when:

• A repository being open-sourced has commits attributed to internal @corp.com email addresses
• Contributors want their public github.com noreply email to appear in the OSS history
• DCO sign-off needs to be retroactively appended to historical commits

Do NOT use this skill for general git rebase or author correction tasks unrelated to migration.

Why This Matters for CNCF

CNCF CLOMonitor and OpenSSF Scorecard measure contributor diversity from commit history. Commits showing only [e-mail] obscure the real contributor graph and prevent contributors from receiving proper attribution on their public GitHub profiles.

Approach A (Vionix's choice): squash to milestone commits — history is minimal, no rewrite needed. Approach B (this skill): preserve full commit history but rewrite author metadata only.

Tools Required

pip install git-filter-repo   # replaces deprecated git filter-branch

Step 1 — Discover Internal Authors

# Extract all unique author name+email pairs from the source repo
git log --format="%an|%ae" | sort -u

Step 2 — Resolve Public GitHub Identities

For each internal author, use the available MCP tools and gh CLI to find their public github.com handle:

2a. Search GHES for login username

MCP tool: mcp__git_viasat_com-viasat-github__search_users
  query: "<First Last>"
→ returns GHES login (e.g. jmaze, mpanjabi, dlgarcia)

2b. Look up corporate AD info via VICE

MCP tool: mcp__viasat-vice__vice_get_user_ad_info_by_email
  email: "[e-mail]"
  full: true
→ confirms firstName, lastName, itAdUser

2c. Search github.com by name

gh api "search/users?q=<First+Last>&per_page=5"
# → compare login, name, company fields

2d. Check public org membership

gh api "orgs/<OrgName>/members?per_page=100" | python3 -c "
import json,sys
[print(m['login']) for m in json.load(sys.stdin)]
"

2e. Fetch public profile to confirm and get noreply email

gh api "users/<login>" | python3 -c "
import json,sys
u=json.load(sys.stdin)
print(u.get('name'), u.get('company'))
print(str(u['id'])+'+'+u['login']+'@users.noreply.github.com')
"

The GitHub noreply format is: <ID>+<login>@users.noreply.github.com Use this as the canonical public email — it never exposes a personal address.

Step 3 — Build the .mailmap File

Format: Canonical Name <public@email> Commit Name <[e-mail]>

# Verified contributors
Julian Maze <[e-mail]> Julian Maze <[e-mail]>
Julian Maze <[e-mail]> Maze, Julian <[e-mail]>

# Unverified — placeholder until contributor confirms their handle
Kevin Yang <[e-mail]> Kevin Yang <[e-mail]>
# TODO: Kevin Yang must provide github.com username

# Service accounts
Vionix Bot <[e-mail]> svc_vionix <[e-mail]>

Commit name variants to cover per person:

• Firstname Lastname
• Lastname, Firstname (Microsoft/LDAP display name format)
• Short login alias (e.g. mdesales)

See .mailmap in this repo for the Vionix contributor map.

Step 4 — Rewrite History with git-filter-repo

Warning: this rewrites all commit SHAs. Coordinate with all collaborators before running.

# Dry-run first (prints what would change, no modification)
git filter-repo --mailmap .mailmap --dry-run

# Apply — rewrites the working repo in place
git filter-repo --mailmap .mailmap

# Optional: retroactively append DCO Signed-off-by to every commit
git filter-repo --commit-callback '
    if b"Signed-off-by" not in commit.message:
        commit.message = commit.message.rstrip() + (
            b"\n\nSigned-off-by: "
            + commit.author_name
            + b" <"
            + commit.author_email
            + b">"
        )
'

Step 5 — Verify

# Confirm no internal emails remain in log
git log --format="%ae" | grep -E "@corp\.com|@internal\." | wc -l
# Expected: 0

# Check shortlog shows canonical names
git shortlog --summary --email | head -20

Resolution Status for This Repo (.mailmap)

Contributor	GHES login	github.com login	Status
Marcello DeSales	mdesales	marcellodesales	✅ Verified
Julian Maze	jmaze	julianmaze	✅ Verified (company=Viasat)
Mehk Panjabi	mpanjabi	mpanjabi	✅ Verified (name match)
David Garcia	dlgarcia	dlgarcia808	⚠️ Org member, name unconfirmed
Kevin Yang	KYANG / kyang	unknown	❌ TODO
James Maier	jamaier	unknown	❌ TODO

Related Skills

• cncf-migration-secrets-scrubber — strip internal content before migration
• cncf-migration-ghes-to-github — push clean history to public repo