Post Merge

post-merge

Read AGENTS.md, then read .agent-teamflow from the repo root, then follow the workflow below. Treat the user's remaining request text as $ARGUMENTS.

---

After merging an integration→staging MR/PR, label its linked issues with the configured "done in staging" label so the issue tracker shows the work has reached staging. Issues stay open until staging→main triggers the tracker's native auto-close.

Run this workflow in an isolated worker if the current agent runtime supports one; otherwise run it in the main session. Keep CLI output concise and report only the final summary.

The label name is read from .agent-teamflow at labels.doneInStaging. If unset, default to the string done-in-staging. All references to `` below resolve to that value.

Prerequisite

The MR/PR must already be MERGED. This skill refuses to operate on:

• MRs/PRs still in opened/open state — not merged yet
• MRs/PRs in closed state — closed-without-merging; no code reached staging

Only merged is acceptable. Run this after you manually merge in the UI.

What this command does

1. Find the merged MR/PR to operate on.
2. Parse Closes # references from its description.
3. Ensure the `` label exists in the project (create if missing).
4. Add the label to each linked issue. Idempotent — re-running is safe.
5. Stop. Does NOT close issues. Does NOT merge anything. Does NOT touch branches.

Input

$ARGUMENTS:

• empty → find the most recently merged MR/PR (source=`, target=`) by the current user within the last 6 hours.
• single integer `` → operate on that specific MR/PR (must be merged).

Setup

Read .agent-teamflow from the repo root. Extract issueTracker, project, branches, owners, and labels.doneInStaging (default done-in-staging if missing) — refer to it as ` below. Resolve using the same owner-resolution logic as git-auto-merge` Step 0.

Label spec

• Name: ``
• Color: #36cd96 (green)
• Description: MR/PR resolving this issue has been merged into staging. Will auto-close when staging → main flows.

Check existence first, only create if missing:

# GitLab
glab api projects//labels | grep -q "" || \
glab api projects//labels --method POST \
  -f "name=" \
  -f "color=#36cd96" \
  -f "description=MR resolving this issue has been merged into staging."

# GitHub — uses labels on the repo
gh api repos//labels | grep -q "" || \
gh api repos//labels --method POST \
  -f "name=" \
  -f "color=36cd96" \
  -f "description=PR resolving this issue has been merged into staging."

Execution

Step 1. Find the MR/PR

If $ARGUMENTS empty:

# GitLab
glab mr list --merged --source-branch  --target-branch  --per-page 10 --output json

# GitHub
gh pr list --state merged --head  --base  --json number,title,mergedAt,author

Filter to MRs/PRs merged within the last 6 hours by the current user.

• Zero matches → exit with No recently merged MR/PR found in the last 6 hours.
• One match → use it.
• Multiple matches → ask the user single-select for the user to pick.

If $ARGUMENTS is a number: fetch that specific MR/PR and verify it is merged. Refuse if still open or closed-without-merging.

Step 2. Extract linked ids

From the MR/PR description, grep for Closes #(\d+) (also match Close|Closing|Fixes|Fix|Resolves). Dedupe.

Verify each id exists. Drop any that 404. Never fabricate.

If no valid ids: exit with MR/PR has no parseable linked issues. Nothing to label.

Step 3. Ensure label exists

(See label spec above.)

Step 4. Apply label to each linked issue

# GitLab — additive, does not replace existing labels
glab api -X PUT projects//issues/ -f "add_labels="

# GitHub
gh api repos//issues//labels --method POST -f "labels[]="

Idempotent. Capture each call's status — note failures in the report but continue.

Step 5. Report

MR/PR # () — merged at 

Linked issues labeled ``:
  # 
  # 

Failures (if any):
  # — 

Issues are still OPEN — will auto-close when staging → main flows.

Constraints

• No emojis.
• Never close issues — only label.
• Never merge MRs/PRs — only consume already-merged ones.
• Never delete branches or worktrees.
• If CLI fails with auth/permission errors, report verbatim and stop.

Typical usage sequence

/git-auto-merge     → push, merge into integration branch, open MR/PR
(you)               → review the MR/PR in the UI, merge it
/post-merge         → label all linked issues with 

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: lkim0402
• Source: lkim0402/agent-teamflow
• License: MIT

Install and usage instructions live in the source repository linked above.